git executable is now configurable via .ini files

development.ini

@@ -1,442 +1,447 b''
 ################################################################################
 ################################################################################
 # RhodeCode - Pylons environment configuration                                 #
 #                                                                              # 
 # The %(here)s variable will be replaced with the parent directory of this file#
 ################################################################################
 
 [DEFAULT]
 debug = true
 pdebug = false
 ################################################################################
 ## Uncomment and replace with the address which should receive                ## 
 ## any error reports after application crash                                  ##
 ## Additionally those settings will be used by RhodeCode mailing system       ##
 ################################################################################
 #email_to = admin@localhost
 #error_email_from = paste_error@localhost
 #app_email_from = rhodecode-noreply@localhost
 #error_message =
 #email_prefix = [RhodeCode]
 
 #smtp_server = mail.server.com
 #smtp_username = 
 #smtp_password = 
 #smtp_port = 
 #smtp_use_tls = false
 #smtp_use_ssl = true
 # Specify available auth parameters here (e.g. LOGIN PLAIN CRAM-MD5, etc.)
 #smtp_auth = 
 
 [server:main]
 ## PASTE
 ##nr of threads to spawn
 #threadpool_workers = 5
 
 ##max request before thread respawn
 #threadpool_max_requests = 10
 
 ##option to use threads of process
 #use_threadpool = true
 
 #use = egg:Paste#http
 
 #WAITRESS
 threads = 5
+#100GB
+max_request_body_size = 107374182400
 use = egg:waitress#main
 
 host = 0.0.0.0
 port = 5000
 
 [filter:proxy-prefix]
 # prefix middleware for rc
 use = egg:PasteDeploy#prefix
 prefix = /<your-prefix>
 
 [app:main]
 use = egg:rhodecode
 #filter-with = proxy-prefix
 full_stack = true
 static_files = true
 # Optional Languages
 # en, fr, ja, pt_BR, zh_CN, zh_TW, pl
 lang = en
 cache_dir = %(here)s/data
 index_dir = %(here)s/data/index
 app_instance_uuid = rc-develop
 cut_off_limit = 256000
 vcs_full_cache = True
 # force https in RhodeCode, fixes https redirects, assumes it's always https
 force_https = false
 # use Strict-Transport-Security headers
 use_htsts = false
 commit_parse_limit = 25
 # number of items displayed in lightweight dashboard before paginating
 dashboard_items = 100
 use_gravatar = true
 
+# path to git executable
+git_path = git
+
 ## RSS feed options
 
 rss_cut_off_limit = 256000
 rss_items_per_page = 10
 rss_include_diff = false
 
 
 ## alternative_gravatar_url allows you to use your own avatar server application
 ## the following parts of the URL will be replaced
 ## {email}        user email
 ## {md5email}     md5 hash of the user email (like at gravatar.com)
 ## {size}         size of the image that is expected from the server application
 ## {scheme}       http/https from RhodeCode server
 ## {netloc}       network location from RhodeCode server
 #alternative_gravatar_url = http://myavatarserver.com/getbyemail/{email}/{size}
 #alternative_gravatar_url = http://myavatarserver.com/getbymd5/{md5email}?s={size}
 
 container_auth_enabled = false
 proxypass_auth_enabled = false
 ## default encoding used to convert from and to unicode
 ## can be also a comma seperated list of encoding in case of mixed encodings
 default_encoding = utf8
 
 ## overwrite schema of clone url
 ## available vars:
 ## scheme - http/https
 ## user - current user
 ## pass - password 
 ## netloc - network location
 ## path - usually repo_name
 
 #clone_uri = {scheme}://{user}{pass}{netloc}{path}
 
 ## issue tracking mapping for commits messages
 ## comment out issue_pat, issue_server, issue_prefix to enable
 
 ## pattern to get the issues from commit messages
 ## default one used here is #<numbers> with a regex passive group for `#`
 ## {id} will be all groups matched from this pattern
 
 issue_pat = (?:\s*#)(\d+)
 
 ## server url to the issue, each {id} will be replaced with match
 ## fetched from the regex and {repo} is replaced with full repository name
 ## including groups {repo_name} is replaced with just name of repo
 
 issue_server_link = https://myissueserver.com/{repo}/issue/{id}
 
 ## prefix to add to link to indicate it's an url
 ## #314 will be replaced by <issue_prefix><id>
 
 issue_prefix = #
 
 ## issue_pat, issue_server_link, issue_prefix can have suffixes to specify
 ## multiple patterns, to other issues server, wiki or others
 ## below an example how to create a wiki pattern 
 #  #wiki-some-id -> https://mywiki.com/some-id
 
 #issue_pat_wiki = (?:wiki-)(.+)
 #issue_server_link_wiki = https://mywiki.com/{id}
 #issue_prefix_wiki = WIKI-
 
 
 ## instance-id prefix
 ## a prefix key for this instance used for cache invalidation when running 
 ## multiple instances of rhodecode, make sure it's globally unique for 
 ## all running rhodecode instances. Leave empty if you don't use it
 instance_id = 
 
 ## alternative return HTTP header for failed authentication. Default HTTP
 ## response is 401 HTTPUnauthorized. Currently HG clients have troubles with 
 ## handling that. Set this variable to 403 to return HTTPForbidden
 auth_ret_code =
 
 ####################################
 ###        CELERY CONFIG        ####
 ####################################
 use_celery = false
 broker.host = localhost
 broker.vhost = rabbitmqhost
 broker.port = 5672
 broker.user = rabbitmq
 broker.password = qweqwe
 
 celery.imports = rhodecode.lib.celerylib.tasks
 
 celery.result.backend = amqp
 celery.result.dburi = amqp://
 celery.result.serialier = json
 
 #celery.send.task.error.emails = true
 #celery.amqp.task.result.expires = 18000
 
 celeryd.concurrency = 2
 #celeryd.log.file = celeryd.log
 celeryd.log.level = debug
 celeryd.max.tasks.per.child = 1
 
 #tasks will never be sent to the queue, but executed locally instead.
 celery.always.eager = false
 
 ####################################
 ###         BEAKER CACHE        ####
 ####################################
 beaker.cache.data_dir=%(here)s/data/cache/data
 beaker.cache.lock_dir=%(here)s/data/cache/lock
 
 beaker.cache.regions=super_short_term,short_term,long_term,sql_cache_short,sql_cache_med,sql_cache_long
 
 beaker.cache.super_short_term.type=memory
 beaker.cache.super_short_term.expire=10
 beaker.cache.super_short_term.key_length = 256
 
 beaker.cache.short_term.type=memory
 beaker.cache.short_term.expire=60
 beaker.cache.short_term.key_length = 256
 
 beaker.cache.long_term.type=memory
 beaker.cache.long_term.expire=36000
 beaker.cache.long_term.key_length = 256
 
 beaker.cache.sql_cache_short.type=memory
 beaker.cache.sql_cache_short.expire=10
 beaker.cache.sql_cache_short.key_length = 256
 
 beaker.cache.sql_cache_med.type=memory
 beaker.cache.sql_cache_med.expire=360
 beaker.cache.sql_cache_med.key_length = 256
 
 beaker.cache.sql_cache_long.type=file
 beaker.cache.sql_cache_long.expire=3600
 beaker.cache.sql_cache_long.key_length = 256
 
 ####################################
 ###       BEAKER SESSION        ####
 ####################################
 ## Type of storage used for the session, current types are 
 ## dbm, file, memcached, database, and memory. 
 ## The storage uses the Container API 
 ## that is also used by the cache system.
 
 ## db session ##
 #beaker.session.type = ext:database
 #beaker.session.sa.url = postgresql://postgres:qwe@localhost/rhodecode
 #beaker.session.table_name = db_session 
 
 ## encrypted cookie client side session, good for many instances ##
 #beaker.session.type = cookie
 
 ## file based cookies (default) ##
 #beaker.session.type = file
 
 
 beaker.session.key = rhodecode
 ## secure cookie requires AES python libraries ##
 #beaker.session.encrypt_key = g654dcno0-9873jhgfreyu
 #beaker.session.validate_key = 9712sds2212c--zxc123
 ## sets session as invalid if it haven't been accessed for given amount of time
 beaker.session.timeout = 2592000
 beaker.session.httponly = true
 #beaker.session.cookie_path = /<your-prefix>
 
 ## uncomment for https secure cookie ##
 beaker.session.secure = false
 
 ## auto save the session to not to use .save() ##
 beaker.session.auto = False
 
 ## default cookie expiration time in seconds `true` expire at browser close ##
 #beaker.session.cookie_expires = 3600
 
 
 ############################
 ## ERROR HANDLING SYSTEMS ##
 ############################
 
 ####################
 ### [errormator] ###
 ####################
 
 # Errormator is tailored to work with RhodeCode, see 
 # http://errormator.com for details how to obtain an account
 # you must install python package `errormator_client` to make it work
 
 # errormator enabled
 errormator = true
 
 errormator.server_url = https://api.errormator.com
 errormator.api_key = YOUR_API_KEY
 
 # TWEAK AMOUNT OF INFO SENT HERE
 
 # enables 404 error logging (default False)
 errormator.report_404 = false
 
 # time in seconds after request is considered being slow (default 1)
 errormator.slow_request_time = 1
 
 # record slow requests in application
 # (needs to be enabled for slow datastore recording and time tracking)
 errormator.slow_requests = true
 
 # enable hooking to application loggers
 # errormator.logging = true
 
 # minimum log level for log capture
 # errormator.logging.level = WARNING
 
 # send logs only from erroneous/slow requests
 # (saves API quota for intensive logging)
 errormator.logging_on_error = false
 
 # list of additonal keywords that should be grabbed from environ object 
 # can be string with comma separated list of words in lowercase
 # (by default client will always send following info:
 # 'REMOTE_USER', 'REMOTE_ADDR', 'SERVER_NAME', 'CONTENT_TYPE' + all keys that 
 # start with HTTP* this list be extended with additional keywords here
 errormator.environ_keys_whitelist = 
 
 
 # list of keywords that should be blanked from request object 
 # can be string with comma separated list of words in lowercase
 # (by default client will always blank keys that contain following words 
 # 'password', 'passwd', 'pwd', 'auth_tkt', 'secret', 'csrf'
 # this list be extended with additional keywords set here
 errormator.request_keys_blacklist =
 
 
 # list of namespaces that should be ignores when gathering log entries
 # can be string with comma separated list of namespaces
 # (by default the client ignores own entries: errormator_client.client)
 errormator.log_namespace_blacklist =  
 
 
 ################
 ### [sentry] ###
 ################
 
 # sentry is a alternative open source error aggregator
 # you must install python packages `sentry` and `raven` to enable 
 
 sentry.dsn = YOUR_DNS
 sentry.servers =
 sentry.name =
 sentry.key =
 sentry.public_key =
 sentry.secret_key =
 sentry.project =
 sentry.site =
 sentry.include_paths =
 sentry.exclude_paths =
 
 
 ################################################################################
 ## WARNING: *THE LINE BELOW MUST BE UNCOMMENTED ON A PRODUCTION ENVIRONMENT*  ##
 ## Debug mode will enable the interactive debugging tool, allowing ANYONE to  ##
 ## execute malicious code after an exception is raised.                       ##
 ################################################################################
 #set debug = false
 
 ##################################
 ###       LOGVIEW CONFIG       ###
 ##################################
 logview.sqlalchemy = #faa
 logview.pylons.templating = #bfb
 logview.pylons.util = #eee
 
 #########################################################
 ### DB CONFIGS - EACH DB WILL HAVE IT'S OWN CONFIG    ###
 #########################################################
 #sqlalchemy.db1.url = sqlite:///%(here)s/rhodecode.db
 sqlalchemy.db1.url = postgresql://postgres:qwe@localhost/rhodecode
 sqlalchemy.db1.echo = false
 sqlalchemy.db1.pool_recycle = 3600
 sqlalchemy.db1.convert_unicode = true
 
 ################################
 ### LOGGING CONFIGURATION   ####
 ################################
 [loggers]
 keys = root, routes, rhodecode, sqlalchemy, beaker, templates, whoosh_indexer
 
 [handlers]
 keys = console, console_sql
 
 [formatters]
 keys = generic, color_formatter, color_formatter_sql
 
 #############
 ## LOGGERS ##
 #############
 [logger_root]
 level = NOTSET
 handlers = console
 
 [logger_routes]
 level = DEBUG
 handlers = 
 qualname = routes.middleware
 # "level = DEBUG" logs the route matched and routing variables.
 propagate = 1
 
 [logger_beaker]
 level = DEBUG
 handlers = 
 qualname = beaker.container
 propagate = 1
 
 [logger_templates]
 level = INFO
 handlers = 
 qualname = pylons.templating
 propagate = 1
 
 [logger_rhodecode]
 level = DEBUG
 handlers = 
 qualname = rhodecode
 propagate = 1
 
 [logger_sqlalchemy]
 level = INFO
 handlers = console_sql
 qualname = sqlalchemy.engine
 propagate = 0
 
 [logger_whoosh_indexer]
 level = DEBUG
 handlers = 
 qualname = whoosh_indexer
 propagate = 1
 
 ##############
 ## HANDLERS ##
 ##############
 
 [handler_console]
 class = StreamHandler
 args = (sys.stderr,)
 level = DEBUG
 formatter = color_formatter
 
 [handler_console_sql]
 class = StreamHandler
 args = (sys.stderr,)
 level = DEBUG
 formatter = color_formatter_sql
 
 ################
 ## FORMATTERS ##
 ################
 
 [formatter_generic]
 format = %(asctime)s.%(msecs)03d %(levelname)-5.5s [%(name)s] %(message)s
 datefmt = %Y-%m-%d %H:%M:%S
 
 [formatter_color_formatter]
 class=rhodecode.lib.colored_formatter.ColorFormatter
 format= %(asctime)s.%(msecs)03d %(levelname)-5.5s [%(name)s] %(message)s
 datefmt = %Y-%m-%d %H:%M:%S
 
 [formatter_color_formatter_sql]
 class=rhodecode.lib.colored_formatter.ColorFormatterSql
 format= %(asctime)s.%(msecs)03d %(levelname)-5.5s [%(name)s] %(message)s
 datefmt = %Y-%m-%d %H:%M:%S

production.ini

@@ -1,442 +1,447 b''
 ################################################################################
 ################################################################################
 # RhodeCode - Pylons environment configuration                                 #
 #                                                                              # 
 # The %(here)s variable will be replaced with the parent directory of this file#
 ################################################################################
 
 [DEFAULT]
 debug = true
 pdebug = false
 ################################################################################
 ## Uncomment and replace with the address which should receive                ## 
 ## any error reports after application crash                                  ##
 ## Additionally those settings will be used by RhodeCode mailing system       ##
 ################################################################################
 #email_to = admin@localhost
 #error_email_from = paste_error@localhost
 #app_email_from = rhodecode-noreply@localhost
 #error_message =
 #email_prefix = [RhodeCode]
 
 #smtp_server = mail.server.com
 #smtp_username = 
 #smtp_password = 
 #smtp_port = 
 #smtp_use_tls = false
 #smtp_use_ssl = true
 # Specify available auth parameters here (e.g. LOGIN PLAIN CRAM-MD5, etc.)
 #smtp_auth = 
 
 [server:main]
 ## PASTE
 ##nr of threads to spawn
 #threadpool_workers = 5
 
 ##max request before thread respawn
 #threadpool_max_requests = 10
 
 ##option to use threads of process
 #use_threadpool = true
 
 #use = egg:Paste#http
 
 #WAITRESS
 threads = 5
+#100GB
+max_request_body_size = 107374182400
 use = egg:waitress#main
 
 host = 127.0.0.1
 port = 8001
 
 [filter:proxy-prefix]
 # prefix middleware for rc
 use = egg:PasteDeploy#prefix
 prefix = /<your-prefix>
 
 [app:main]
 use = egg:rhodecode
 #filter-with = proxy-prefix
 full_stack = true
 static_files = true
 # Optional Languages
 # en, fr, ja, pt_BR, zh_CN, zh_TW, pl
 lang = en
 cache_dir = %(here)s/data
 index_dir = %(here)s/data/index
 app_instance_uuid = rc-production
 cut_off_limit = 256000
 vcs_full_cache = True
 # force https in RhodeCode, fixes https redirects, assumes it's always https
 force_https = false
 # use Strict-Transport-Security headers
 use_htsts = false
 commit_parse_limit = 50
 # number of items displayed in lightweight dashboard before paginating
 dashboard_items = 100
 use_gravatar = true
 
+# path to git executable
+git_path = git
+
 ## RSS feed options
 
 rss_cut_off_limit = 256000
 rss_items_per_page = 10
 rss_include_diff = false
 
 
 ## alternative_gravatar_url allows you to use your own avatar server application
 ## the following parts of the URL will be replaced
 ## {email}        user email
 ## {md5email}     md5 hash of the user email (like at gravatar.com)
 ## {size}         size of the image that is expected from the server application
 ## {scheme}       http/https from RhodeCode server
 ## {netloc}       network location from RhodeCode server
 #alternative_gravatar_url = http://myavatarserver.com/getbyemail/{email}/{size}
 #alternative_gravatar_url = http://myavatarserver.com/getbymd5/{md5email}?s={size}
 
 container_auth_enabled = false
 proxypass_auth_enabled = false
 ## default encoding used to convert from and to unicode
 ## can be also a comma seperated list of encoding in case of mixed encodings
 default_encoding = utf8
 
 ## overwrite schema of clone url
 ## available vars:
 ## scheme - http/https
 ## user - current user
 ## pass - password 
 ## netloc - network location
 ## path - usually repo_name
 
 #clone_uri = {scheme}://{user}{pass}{netloc}{path}
 
 ## issue tracking mapping for commits messages
 ## comment out issue_pat, issue_server, issue_prefix to enable
 
 ## pattern to get the issues from commit messages
 ## default one used here is #<numbers> with a regex passive group for `#`
 ## {id} will be all groups matched from this pattern
 
 issue_pat = (?:\s*#)(\d+)
 
 ## server url to the issue, each {id} will be replaced with match
 ## fetched from the regex and {repo} is replaced with full repository name
 ## including groups {repo_name} is replaced with just name of repo
 
 issue_server_link = https://myissueserver.com/{repo}/issue/{id}
 
 ## prefix to add to link to indicate it's an url
 ## #314 will be replaced by <issue_prefix><id>
 
 issue_prefix = #
 
 ## issue_pat, issue_server_link, issue_prefix can have suffixes to specify
 ## multiple patterns, to other issues server, wiki or others
 ## below an example how to create a wiki pattern 
 #  #wiki-some-id -> https://mywiki.com/some-id
 
 #issue_pat_wiki = (?:wiki-)(.+)
 #issue_server_link_wiki = https://mywiki.com/{id}
 #issue_prefix_wiki = WIKI-
 
 
 ## instance-id prefix
 ## a prefix key for this instance used for cache invalidation when running 
 ## multiple instances of rhodecode, make sure it's globally unique for 
 ## all running rhodecode instances. Leave empty if you don't use it
 instance_id = 
 
 ## alternative return HTTP header for failed authentication. Default HTTP
 ## response is 401 HTTPUnauthorized. Currently HG clients have troubles with 
 ## handling that. Set this variable to 403 to return HTTPForbidden
 auth_ret_code =
 
 ####################################
 ###        CELERY CONFIG        ####
 ####################################
 use_celery = false
 broker.host = localhost
 broker.vhost = rabbitmqhost
 broker.port = 5672
 broker.user = rabbitmq
 broker.password = qweqwe
 
 celery.imports = rhodecode.lib.celerylib.tasks
 
 celery.result.backend = amqp
 celery.result.dburi = amqp://
 celery.result.serialier = json
 
 #celery.send.task.error.emails = true
 #celery.amqp.task.result.expires = 18000
 
 celeryd.concurrency = 2
 #celeryd.log.file = celeryd.log
 celeryd.log.level = debug
 celeryd.max.tasks.per.child = 1
 
 #tasks will never be sent to the queue, but executed locally instead.
 celery.always.eager = false
 
 ####################################
 ###         BEAKER CACHE        ####
 ####################################
 beaker.cache.data_dir=%(here)s/data/cache/data
 beaker.cache.lock_dir=%(here)s/data/cache/lock
 
 beaker.cache.regions=super_short_term,short_term,long_term,sql_cache_short,sql_cache_med,sql_cache_long
 
 beaker.cache.super_short_term.type=memory
 beaker.cache.super_short_term.expire=10
 beaker.cache.super_short_term.key_length = 256
 
 beaker.cache.short_term.type=memory
 beaker.cache.short_term.expire=60
 beaker.cache.short_term.key_length = 256
 
 beaker.cache.long_term.type=memory
 beaker.cache.long_term.expire=36000
 beaker.cache.long_term.key_length = 256
 
 beaker.cache.sql_cache_short.type=memory
 beaker.cache.sql_cache_short.expire=10
 beaker.cache.sql_cache_short.key_length = 256
 
 beaker.cache.sql_cache_med.type=memory
 beaker.cache.sql_cache_med.expire=360
 beaker.cache.sql_cache_med.key_length = 256
 
 beaker.cache.sql_cache_long.type=file
 beaker.cache.sql_cache_long.expire=3600
 beaker.cache.sql_cache_long.key_length = 256
 
 ####################################
 ###       BEAKER SESSION        ####
 ####################################
 ## Type of storage used for the session, current types are 
 ## dbm, file, memcached, database, and memory. 
 ## The storage uses the Container API 
 ## that is also used by the cache system.
 
 ## db session ##
 #beaker.session.type = ext:database
 #beaker.session.sa.url = postgresql://postgres:qwe@localhost/rhodecode
 #beaker.session.table_name = db_session 
 
 ## encrypted cookie client side session, good for many instances ##
 #beaker.session.type = cookie
 
 ## file based cookies (default) ##
 #beaker.session.type = file
 
 
 beaker.session.key = rhodecode
 ## secure cookie requires AES python libraries ##
 #beaker.session.encrypt_key = g654dcno0-9873jhgfreyu
 #beaker.session.validate_key = 9712sds2212c--zxc123
 ## sets session as invalid if it haven't been accessed for given amount of time
 beaker.session.timeout = 2592000
 beaker.session.httponly = true
 #beaker.session.cookie_path = /<your-prefix>
 
 ## uncomment for https secure cookie ##
 beaker.session.secure = false
 
 ## auto save the session to not to use .save() ##
 beaker.session.auto = False
 
 ## default cookie expiration time in seconds `true` expire at browser close ##
 #beaker.session.cookie_expires = 3600
 
 
 ############################
 ## ERROR HANDLING SYSTEMS ##
 ############################
 
 ####################
 ### [errormator] ###
 ####################
 
 # Errormator is tailored to work with RhodeCode, see 
 # http://errormator.com for details how to obtain an account
 # you must install python package `errormator_client` to make it work
 
 # errormator enabled
 errormator = true
 
 errormator.server_url = https://api.errormator.com
 errormator.api_key = YOUR_API_KEY
 
 # TWEAK AMOUNT OF INFO SENT HERE
 
 # enables 404 error logging (default False)
 errormator.report_404 = false
 
 # time in seconds after request is considered being slow (default 1)
 errormator.slow_request_time = 1
 
 # record slow requests in application
 # (needs to be enabled for slow datastore recording and time tracking)
 errormator.slow_requests = true
 
 # enable hooking to application loggers
 # errormator.logging = true
 
 # minimum log level for log capture
 # errormator.logging.level = WARNING
 
 # send logs only from erroneous/slow requests
 # (saves API quota for intensive logging)
 errormator.logging_on_error = false
 
 # list of additonal keywords that should be grabbed from environ object 
 # can be string with comma separated list of words in lowercase
 # (by default client will always send following info:
 # 'REMOTE_USER', 'REMOTE_ADDR', 'SERVER_NAME', 'CONTENT_TYPE' + all keys that 
 # start with HTTP* this list be extended with additional keywords here
 errormator.environ_keys_whitelist = 
 
 
 # list of keywords that should be blanked from request object 
 # can be string with comma separated list of words in lowercase
 # (by default client will always blank keys that contain following words 
 # 'password', 'passwd', 'pwd', 'auth_tkt', 'secret', 'csrf'
 # this list be extended with additional keywords set here
 errormator.request_keys_blacklist =
 
 
 # list of namespaces that should be ignores when gathering log entries
 # can be string with comma separated list of namespaces
 # (by default the client ignores own entries: errormator_client.client)
 errormator.log_namespace_blacklist =  
 
 
 ################
 ### [sentry] ###
 ################
 
 # sentry is a alternative open source error aggregator
 # you must install python packages `sentry` and `raven` to enable 
 
 sentry.dsn = YOUR_DNS
 sentry.servers =
 sentry.name =
 sentry.key =
 sentry.public_key =
 sentry.secret_key =
 sentry.project =
 sentry.site =
 sentry.include_paths =
 sentry.exclude_paths =
 
 
 ################################################################################
 ## WARNING: *THE LINE BELOW MUST BE UNCOMMENTED ON A PRODUCTION ENVIRONMENT*  ##
 ## Debug mode will enable the interactive debugging tool, allowing ANYONE to  ##
 ## execute malicious code after an exception is raised.                       ##
 ################################################################################
 set debug = false
 
 ##################################
 ###       LOGVIEW CONFIG       ###
 ##################################
 logview.sqlalchemy = #faa
 logview.pylons.templating = #bfb
 logview.pylons.util = #eee
 
 #########################################################
 ### DB CONFIGS - EACH DB WILL HAVE IT'S OWN CONFIG    ###
 #########################################################
 #sqlalchemy.db1.url = sqlite:///%(here)s/rhodecode.db
 sqlalchemy.db1.url = postgresql://postgres:qwe@localhost/rhodecode
 sqlalchemy.db1.echo = false
 sqlalchemy.db1.pool_recycle = 3600
 sqlalchemy.db1.convert_unicode = true
 
 ################################
 ### LOGGING CONFIGURATION   ####
 ################################
 [loggers]
 keys = root, routes, rhodecode, sqlalchemy, beaker, templates, whoosh_indexer
 
 [handlers]
 keys = console, console_sql
 
 [formatters]
 keys = generic, color_formatter, color_formatter_sql
 
 #############
 ## LOGGERS ##
 #############
 [logger_root]
 level = NOTSET
 handlers = console
 
 [logger_routes]
 level = DEBUG
 handlers = 
 qualname = routes.middleware
 # "level = DEBUG" logs the route matched and routing variables.
 propagate = 1
 
 [logger_beaker]
 level = DEBUG
 handlers = 
 qualname = beaker.container
 propagate = 1
 
 [logger_templates]
 level = INFO
 handlers = 
 qualname = pylons.templating
 propagate = 1
 
 [logger_rhodecode]
 level = DEBUG
 handlers = 
 qualname = rhodecode
 propagate = 1
 
 [logger_sqlalchemy]
 level = INFO
 handlers = console_sql
 qualname = sqlalchemy.engine
 propagate = 0
 
 [logger_whoosh_indexer]
 level = DEBUG
 handlers = 
 qualname = whoosh_indexer
 propagate = 1
 
 ##############
 ## HANDLERS ##
 ##############
 
 [handler_console]
 class = StreamHandler
 args = (sys.stderr,)
 level = INFO
 formatter = generic
 
 [handler_console_sql]
 class = StreamHandler
 args = (sys.stderr,)
 level = WARN
 formatter = generic
 
 ################
 ## FORMATTERS ##
 ################
 
 [formatter_generic]
 format = %(asctime)s.%(msecs)03d %(levelname)-5.5s [%(name)s] %(message)s
 datefmt = %Y-%m-%d %H:%M:%S
 
 [formatter_color_formatter]
 class=rhodecode.lib.colored_formatter.ColorFormatter
 format= %(asctime)s.%(msecs)03d %(levelname)-5.5s [%(name)s] %(message)s
 datefmt = %Y-%m-%d %H:%M:%S
 
 [formatter_color_formatter_sql]
 class=rhodecode.lib.colored_formatter.ColorFormatterSql
 format= %(asctime)s.%(msecs)03d %(levelname)-5.5s [%(name)s] %(message)s
 datefmt = %Y-%m-%d %H:%M:%S

rhodecode/config/deployment.ini_tmpl

@@ -1,452 +1,457 b''
 ################################################################################
 ################################################################################
 # RhodeCode - Pylons environment configuration                                 #
 #                                                                              # 
 # The %(here)s variable will be replaced with the parent directory of this file#
 ################################################################################
 
 [DEFAULT]
 debug = true
 pdebug = false
 ################################################################################
 ## Uncomment and replace with the address which should receive                ## 
 ## any error reports after application crash                                  ##
 ## Additionally those settings will be used by RhodeCode mailing system       ##
 ################################################################################
 #email_to = admin@localhost
 #error_email_from = paste_error@localhost
 #app_email_from = rhodecode-noreply@localhost
 #error_message =
 #email_prefix = [RhodeCode]
 
 #smtp_server = mail.server.com
 #smtp_username = 
 #smtp_password = 
 #smtp_port = 
 #smtp_use_tls = false
 #smtp_use_ssl = true
 # Specify available auth parameters here (e.g. LOGIN PLAIN CRAM-MD5, etc.)
 #smtp_auth = 
 
 [server:main]
 ## PASTE
 ##nr of threads to spawn
 #threadpool_workers = 5
 
 ##max request before thread respawn
 #threadpool_max_requests = 10
 
 ##option to use threads of process
 #use_threadpool = true
 
 #use = egg:Paste#http
 
 #WAITRESS
 threads = 5
+#100GB
+max_request_body_size = 107374182400
 use = egg:waitress#main
 
 host = 127.0.0.1
 port = 5000
 
 [filter:proxy-prefix]
 # prefix middleware for rc
 use = egg:PasteDeploy#prefix
 prefix = /<your-prefix>
 
 [app:main]
 use = egg:rhodecode
 #filter-with = proxy-prefix
 full_stack = true
 static_files = true
 # Optional Languages
 # en, fr, ja, pt_BR, zh_CN, zh_TW, pl
 lang = en
 cache_dir = %(here)s/data
 index_dir = %(here)s/data/index
 app_instance_uuid = ${app_instance_uuid}
 cut_off_limit = 256000
 vcs_full_cache = True
 # force https in RhodeCode, fixes https redirects, assumes it's always https
 force_https = false
 # use Strict-Transport-Security headers
 use_htsts = false
 commit_parse_limit = 50
 # number of items displayed in lightweight dashboard before paginating
 dashboard_items = 100
 use_gravatar = true
 
+# path to git executable
+git_path = git
+
 ## RSS feed options
 
 rss_cut_off_limit = 256000
 rss_items_per_page = 10
 rss_include_diff = false
 
 
 ## alternative_gravatar_url allows you to use your own avatar server application
 ## the following parts of the URL will be replaced
 ## {email}        user email
 ## {md5email}     md5 hash of the user email (like at gravatar.com)
 ## {size}         size of the image that is expected from the server application
 ## {scheme}       http/https from RhodeCode server
 ## {netloc}       network location from RhodeCode server
 #alternative_gravatar_url = http://myavatarserver.com/getbyemail/{email}/{size}
 #alternative_gravatar_url = http://myavatarserver.com/getbymd5/{md5email}?s={size}
 
 container_auth_enabled = false
 proxypass_auth_enabled = false
 ## default encoding used to convert from and to unicode
 ## can be also a comma seperated list of encoding in case of mixed encodings
 default_encoding = utf8
 
 ## overwrite schema of clone url
 ## available vars:
 ## scheme - http/https
 ## user - current user
 ## pass - password 
 ## netloc - network location
 ## path - usually repo_name
 
 #clone_uri = {scheme}://{user}{pass}{netloc}{path}
 
 ## issue tracking mapping for commits messages
 ## comment out issue_pat, issue_server, issue_prefix to enable
 
 ## pattern to get the issues from commit messages
 ## default one used here is #<numbers> with a regex passive group for `#`
 ## {id} will be all groups matched from this pattern
 
 issue_pat = (?:\s*#)(\d+)
 
 ## server url to the issue, each {id} will be replaced with match
 ## fetched from the regex and {repo} is replaced with full repository name
 ## including groups {repo_name} is replaced with just name of repo
 
 issue_server_link = https://myissueserver.com/{repo}/issue/{id}
 
 ## prefix to add to link to indicate it's an url
 ## #314 will be replaced by <issue_prefix><id>
 
 issue_prefix = #
 
 ## issue_pat, issue_server_link, issue_prefix can have suffixes to specify
 ## multiple patterns, to other issues server, wiki or others
 ## below an example how to create a wiki pattern 
 #  #wiki-some-id -> https://mywiki.com/some-id
 
 #issue_pat_wiki = (?:wiki-)(.+)
 #issue_server_link_wiki = https://mywiki.com/{id}
 #issue_prefix_wiki = WIKI-
 
 
 ## instance-id prefix
 ## a prefix key for this instance used for cache invalidation when running 
 ## multiple instances of rhodecode, make sure it's globally unique for 
 ## all running rhodecode instances. Leave empty if you don't use it
 instance_id = 
 
 ## alternative return HTTP header for failed authentication. Default HTTP
 ## response is 401 HTTPUnauthorized. Currently HG clients have troubles with 
 ## handling that. Set this variable to 403 to return HTTPForbidden
 auth_ret_code =
 
 ####################################
 ###        CELERY CONFIG        ####
 ####################################
 use_celery = false
 broker.host = localhost
 broker.vhost = rabbitmqhost
 broker.port = 5672
 broker.user = rabbitmq
 broker.password = qweqwe
 
 celery.imports = rhodecode.lib.celerylib.tasks
 
 celery.result.backend = amqp
 celery.result.dburi = amqp://
 celery.result.serialier = json
 
 #celery.send.task.error.emails = true
 #celery.amqp.task.result.expires = 18000
 
 celeryd.concurrency = 2
 #celeryd.log.file = celeryd.log
 celeryd.log.level = debug
 celeryd.max.tasks.per.child = 1
 
 #tasks will never be sent to the queue, but executed locally instead.
 celery.always.eager = false
 
 ####################################
 ###         BEAKER CACHE        ####
 ####################################
 beaker.cache.data_dir=%(here)s/data/cache/data
 beaker.cache.lock_dir=%(here)s/data/cache/lock
 
 beaker.cache.regions=super_short_term,short_term,long_term,sql_cache_short,sql_cache_med,sql_cache_long
 
 beaker.cache.super_short_term.type=memory
 beaker.cache.super_short_term.expire=10
 beaker.cache.super_short_term.key_length = 256
 
 beaker.cache.short_term.type=memory
 beaker.cache.short_term.expire=60
 beaker.cache.short_term.key_length = 256
 
 beaker.cache.long_term.type=memory
 beaker.cache.long_term.expire=36000
 beaker.cache.long_term.key_length = 256
 
 beaker.cache.sql_cache_short.type=memory
 beaker.cache.sql_cache_short.expire=10
 beaker.cache.sql_cache_short.key_length = 256
 
 beaker.cache.sql_cache_med.type=memory
 beaker.cache.sql_cache_med.expire=360
 beaker.cache.sql_cache_med.key_length = 256
 
 beaker.cache.sql_cache_long.type=file
 beaker.cache.sql_cache_long.expire=3600
 beaker.cache.sql_cache_long.key_length = 256
 
 ####################################
 ###       BEAKER SESSION        ####
 ####################################
 ## Type of storage used for the session, current types are 
 ## dbm, file, memcached, database, and memory. 
 ## The storage uses the Container API 
 ## that is also used by the cache system.
 
 ## db session ##
 #beaker.session.type = ext:database
 #beaker.session.sa.url = postgresql://postgres:qwe@localhost/rhodecode
 #beaker.session.table_name = db_session 
 
 ## encrypted cookie client side session, good for many instances ##
 #beaker.session.type = cookie
 
 ## file based cookies (default) ##
 #beaker.session.type = file
 
 
 beaker.session.key = rhodecode
 ## secure cookie requires AES python libraries ##
 #beaker.session.encrypt_key = g654dcno0-9873jhgfreyu
 #beaker.session.validate_key = 9712sds2212c--zxc123
 ## sets session as invalid if it haven't been accessed for given amount of time
 beaker.session.timeout = 2592000
 beaker.session.httponly = true
 #beaker.session.cookie_path = /<your-prefix>
 
 ## uncomment for https secure cookie ##
 beaker.session.secure = false
 
 ## auto save the session to not to use .save() ##
 beaker.session.auto = False
 
 ## default cookie expiration time in seconds `true` expire at browser close ##
 #beaker.session.cookie_expires = 3600
 
 
 ############################
 ## ERROR HANDLING SYSTEMS ##
 ############################
 
 ####################
 ### [errormator] ###
 ####################
 
 # Errormator is tailored to work with RhodeCode, see 
 # http://errormator.com for details how to obtain an account
 # you must install python package `errormator_client` to make it work
 
 # errormator enabled
 errormator = true
 
 errormator.server_url = https://api.errormator.com
 errormator.api_key = YOUR_API_KEY
 
 # TWEAK AMOUNT OF INFO SENT HERE
 
 # enables 404 error logging (default False)
 errormator.report_404 = false
 
 # time in seconds after request is considered being slow (default 1)
 errormator.slow_request_time = 1
 
 # record slow requests in application
 # (needs to be enabled for slow datastore recording and time tracking)
 errormator.slow_requests = true
 
 # enable hooking to application loggers
 # errormator.logging = true
 
 # minimum log level for log capture
 # errormator.logging.level = WARNING
 
 # send logs only from erroneous/slow requests
 # (saves API quota for intensive logging)
 errormator.logging_on_error = false
 
 # list of additonal keywords that should be grabbed from environ object 
 # can be string with comma separated list of words in lowercase
 # (by default client will always send following info:
 # 'REMOTE_USER', 'REMOTE_ADDR', 'SERVER_NAME', 'CONTENT_TYPE' + all keys that 
 # start with HTTP* this list be extended with additional keywords here
 errormator.environ_keys_whitelist = 
 
 
 # list of keywords that should be blanked from request object 
 # can be string with comma separated list of words in lowercase
 # (by default client will always blank keys that contain following words 
 # 'password', 'passwd', 'pwd', 'auth_tkt', 'secret', 'csrf'
 # this list be extended with additional keywords set here
 errormator.request_keys_blacklist =
 
 
 # list of namespaces that should be ignores when gathering log entries
 # can be string with comma separated list of namespaces
 # (by default the client ignores own entries: errormator_client.client)
 errormator.log_namespace_blacklist =  
 
 
 ################
 ### [sentry] ###
 ################
 
 # sentry is a alternative open source error aggregator
 # you must install python packages `sentry` and `raven` to enable 
 
 sentry.dsn = YOUR_DNS
 sentry.servers =
 sentry.name =
 sentry.key =
 sentry.public_key =
 sentry.secret_key =
 sentry.project =
 sentry.site =
 sentry.include_paths =
 sentry.exclude_paths =
 
 
 ################################################################################
 ## WARNING: *THE LINE BELOW MUST BE UNCOMMENTED ON A PRODUCTION ENVIRONMENT*  ##
 ## Debug mode will enable the interactive debugging tool, allowing ANYONE to  ##
 ## execute malicious code after an exception is raised.                       ##
 ################################################################################
 set debug = false
 
 ##################################
 ###       LOGVIEW CONFIG       ###
 ##################################
 logview.sqlalchemy = #faa
 logview.pylons.templating = #bfb
 logview.pylons.util = #eee
 
 #########################################################
 ### DB CONFIGS - EACH DB WILL HAVE IT'S OWN CONFIG    ###
 #########################################################
 
 # SQLITE [default]
 sqlalchemy.db1.url = sqlite:///%(here)s/rhodecode.db
  
 # POSTGRESQL
 # sqlalchemy.db1.url = postgresql://user:pass@localhost/rhodecode
 
 # MySQL
 # sqlalchemy.db1.url = mysql://user:pass@localhost/rhodecode
 
 # see sqlalchemy docs for others
 
 sqlalchemy.db1.echo = false
 sqlalchemy.db1.pool_recycle = 3600
 sqlalchemy.db1.convert_unicode = true
 
 ################################
 ### LOGGING CONFIGURATION   ####
 ################################
 [loggers]
 keys = root, routes, rhodecode, sqlalchemy, beaker, templates, whoosh_indexer
 
 [handlers]
 keys = console, console_sql
 
 [formatters]
 keys = generic, color_formatter, color_formatter_sql
 
 #############
 ## LOGGERS ##
 #############
 [logger_root]
 level = NOTSET
 handlers = console
 
 [logger_routes]
 level = DEBUG
 handlers = 
 qualname = routes.middleware
 # "level = DEBUG" logs the route matched and routing variables.
 propagate = 1
 
 [logger_beaker]
 level = DEBUG
 handlers = 
 qualname = beaker.container
 propagate = 1
 
 [logger_templates]
 level = INFO
 handlers = 
 qualname = pylons.templating
 propagate = 1
 
 [logger_rhodecode]
 level = DEBUG
 handlers = 
 qualname = rhodecode
 propagate = 1
 
 [logger_sqlalchemy]
 level = INFO
 handlers = console_sql
 qualname = sqlalchemy.engine
 propagate = 0
 
 [logger_whoosh_indexer]
 level = DEBUG
 handlers = 
 qualname = whoosh_indexer
 propagate = 1
 
 ##############
 ## HANDLERS ##
 ##############
 
 [handler_console]
 class = StreamHandler
 args = (sys.stderr,)
 level = INFO
 formatter = generic
 
 [handler_console_sql]
 class = StreamHandler
 args = (sys.stderr,)
 level = WARN
 formatter = generic
 
 ################
 ## FORMATTERS ##
 ################
 
 [formatter_generic]
 format = %(asctime)s.%(msecs)03d %(levelname)-5.5s [%(name)s] %(message)s
 datefmt = %Y-%m-%d %H:%M:%S
 
 [formatter_color_formatter]
 class=rhodecode.lib.colored_formatter.ColorFormatter
 format= %(asctime)s.%(msecs)03d %(levelname)-5.5s [%(name)s] %(message)s
 datefmt = %Y-%m-%d %H:%M:%S
 
 [formatter_color_formatter_sql]
 class=rhodecode.lib.colored_formatter.ColorFormatterSql
 format= %(asctime)s.%(msecs)03d %(levelname)-5.5s [%(name)s] %(message)s
 datefmt = %Y-%m-%d %H:%M:%S

rhodecode/lib/middleware/pygrack.py

@@ -1,200 +1,203 b''
 import os
 import socket
 import logging
 import subprocess
 import traceback
 
 from webob import Request, Response, exc
 
+import rhodecode
 from rhodecode.lib import subprocessio
 
 log = logging.getLogger(__name__)
 
 
 class FileWrapper(object):
 
     def __init__(self, fd, content_length):
         self.fd = fd
         self.content_length = content_length
         self.remain = content_length
 
     def read(self, size):
         if size <= self.remain:
             try:
                 data = self.fd.read(size)
             except socket.error:
                 raise IOError(self)
             self.remain -= size
         elif self.remain:
             data = self.fd.read(self.remain)
             self.remain = 0
         else:
             data = None
         return data
 
     def __repr__(self):
         return '<FileWrapper %s len: %s, read: %s>' % (
             self.fd, self.content_length, self.content_length - self.remain
         )
 
 
 class GitRepository(object):
     git_folder_signature = set(['config', 'head', 'info', 'objects', 'refs'])
     commands = ['git-upload-pack', 'git-receive-pack']
 
     def __init__(self, repo_name, content_path, extras):
         files = set([f.lower() for f in os.listdir(content_path)])
         if  not (self.git_folder_signature.intersection(files)
                 == self.git_folder_signature):
             raise OSError('%s missing git signature' % content_path)
         self.content_path = content_path
         self.valid_accepts = ['application/x-%s-result' %
                               c for c in self.commands]
         self.repo_name = repo_name
         self.extras = extras
 
     def _get_fixedpath(self, path):
         """
         Small fix for repo_path
 
         :param path:
         :type path:
         """
         return path.split(self.repo_name, 1)[-1].strip('/')
 
     def inforefs(self, request, environ):
         """
         WSGI Response producer for HTTP GET Git Smart
         HTTP /info/refs request.
         """
 
         git_command = request.GET.get('service')
         if git_command not in self.commands:
             log.debug('command %s not allowed' % git_command)
             return exc.HTTPMethodNotAllowed()
 
         # note to self:
         # please, resist the urge to add '\n' to git capture and increment
         # line count by 1.
         # The code in Git client not only does NOT need '\n', but actually
         # blows up if you sprinkle "flush" (0000) as "0001\n".
         # It reads binary, per number of bytes specified.
         # if you do add '\n' as part of data, count it.
         server_advert = '# service=%s' % git_command
         packet_len = str(hex(len(server_advert) + 4)[2:].rjust(4, '0')).lower()
+        _git_path = rhodecode.CONFIG.get('git_path', 'git')
         try:
             out = subprocessio.SubprocessIOChunker(
-                r'git %s --stateless-rpc --advertise-refs "%s"' % (
-                                git_command[4:], self.content_path),
+                r'%s %s --stateless-rpc --advertise-refs "%s"' % (
+                            _git_path, git_command[4:], self.content_path),
                 starting_values=[
                     packet_len + server_advert + '0000'
                 ]
             )
         except EnvironmentError, e:
             log.error(traceback.format_exc())
             raise exc.HTTPExpectationFailed()
         resp = Response()
         resp.content_type = 'application/x-%s-advertisement' % str(git_command)
         resp.charset = None
         resp.app_iter = out
         return resp
 
     def backend(self, request, environ):
         """
         WSGI Response producer for HTTP POST Git Smart HTTP requests.
         Reads commands and data from HTTP POST's body.
         returns an iterator obj with contents of git command's
         response to stdout
         """
         git_command = self._get_fixedpath(request.path_info)
         if git_command not in self.commands:
             log.debug('command %s not allowed' % git_command)
             return exc.HTTPMethodNotAllowed()
 
         if 'CONTENT_LENGTH' in environ:
             inputstream = FileWrapper(environ['wsgi.input'],
                                       request.content_length)
         else:
             inputstream = environ['wsgi.input']
 
         try:
             gitenv = os.environ
             from rhodecode.lib.compat import json
             gitenv['RHODECODE_EXTRAS'] = json.dumps(self.extras)
             # forget all configs
             gitenv['GIT_CONFIG_NOGLOBAL'] = '1'
             opts = dict(
                 env=gitenv,
                 cwd=os.getcwd()
             )
             cmd = r'git %s --stateless-rpc "%s"' % (git_command[4:],
                                                     self.content_path),
             log.debug('handling cmd %s' % cmd)
             out = subprocessio.SubprocessIOChunker(
                 cmd,
                 inputstream=inputstream,
                 **opts
             )
         except EnvironmentError, e:
             log.error(traceback.format_exc())
             raise exc.HTTPExpectationFailed()
 
         if git_command in [u'git-receive-pack']:
             # updating refs manually after each push.
             # Needed for pre-1.7.0.4 git clients using regular HTTP mode.
-            cmd = (u'git --git-dir "%s" '
-                    'update-server-info' % self.content_path)
+            _git_path = rhodecode.CONFIG.get('git_path', 'git')
+            cmd = (u'%s --git-dir "%s" '
+                    'update-server-info' % (_git_path, self.content_path))
             log.debug('handling cmd %s' % cmd)
             subprocess.call(cmd, shell=True)
 
         resp = Response()
         resp.content_type = 'application/x-%s-result' % git_command.encode('utf8')
         resp.charset = None
         resp.app_iter = out
         return resp
 
     def __call__(self, environ, start_response):
         request = Request(environ)
         _path = self._get_fixedpath(request.path_info)
         if _path.startswith('info/refs'):
             app = self.inforefs
         elif [a for a in self.valid_accepts if a in request.accept]:
             app = self.backend
         try:
             resp = app(request, environ)
         except exc.HTTPException, e:
             resp = e
             log.error(traceback.format_exc())
         except Exception, e:
             log.error(traceback.format_exc())
             resp = exc.HTTPInternalServerError()
         return resp(environ, start_response)
 
 
 class GitDirectory(object):
 
     def __init__(self, repo_root, repo_name, extras):
         repo_location = os.path.join(repo_root, repo_name)
         if not os.path.isdir(repo_location):
             raise OSError(repo_location)
 
         self.content_path = repo_location
         self.repo_name = repo_name
         self.repo_location = repo_location
         self.extras = extras
 
     def __call__(self, environ, start_response):
         content_path = self.content_path
         try:
             app = GitRepository(self.repo_name, content_path, self.extras)
         except (AssertionError, OSError):
             content_path = os.path.join(content_path, '.git')
             if os.path.isdir(content_path):
                 app = GitRepository(self.repo_name, content_path, self.extras)
             else:
                 return exc.HTTPNotFound()(environ, start_response)
         return app(environ, start_response)
 
 
 def make_wsgi_app(repo_name, repo_root, extras):
     return GitDirectory(repo_root, repo_name, extras)

rhodecode/lib/utils.py

@@ -1,801 +1,800 b''
 # -*- coding: utf-8 -*-
 """
     rhodecode.lib.utils
     ~~~~~~~~~~~~~~~~~~~
 
     Utilities library for RhodeCode
 
     :created_on: Apr 18, 2010
     :author: marcink
     :copyright: (C) 2010-2012 Marcin Kuzminski <[e-mail]>
     :license: GPLv3, see COPYING for more details.
 """
 # This program is free software: you can redistribute it and/or modify
 # it under the terms of the GNU General Public License as published by
 # the Free Software Foundation, either version 3 of the License, or
 # (at your option) any later version.
 #
 # This program is distributed in the hope that it will be useful,
 # but WITHOUT ANY WARRANTY; without even the implied warranty of
 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 # GNU General Public License for more details.
 #
 # You should have received a copy of the GNU General Public License
 # along with this program.  If not, see <http://www.gnu.org/licenses/>.
 
 import os
 import re
 import logging
 import datetime
 import traceback
 import paste
 import beaker
 import tarfile
 import shutil
 import decorator
 import warnings
 from os.path import abspath
 from os.path import dirname as dn, join as jn
 
 from paste.script.command import Command, BadCommand
 
 from mercurial import ui, config
 
 from webhelpers.text import collapse, remove_formatting, strip_tags
 
 from rhodecode.lib.vcs import get_backend
 from rhodecode.lib.vcs.backends.base import BaseChangeset
 from rhodecode.lib.vcs.utils.lazy import LazyProperty
 from rhodecode.lib.vcs.utils.helpers import get_scm
 from rhodecode.lib.vcs.exceptions import VCSError
 
 from rhodecode.lib.caching_query import FromCache
 
 from rhodecode.model import meta
 from rhodecode.model.db import Repository, User, RhodeCodeUi, \
     UserLog, RepoGroup, RhodeCodeSetting, CacheInvalidation
 from rhodecode.model.meta import Session
 from rhodecode.model.repos_group import ReposGroupModel
 from rhodecode.lib.utils2 import safe_str, safe_unicode
 from rhodecode.lib.vcs.utils.fakemod import create_module
 
 log = logging.getLogger(__name__)
 
 REMOVED_REPO_PAT = re.compile(r'rm__\d{8}_\d{6}_\d{6}__.*')
 
 
 def recursive_replace(str_, replace=' '):
     """
     Recursive replace of given sign to just one instance
 
     :param str_: given string
     :param replace: char to find and replace multiple instances
 
     Examples::
     >>> recursive_replace("Mighty---Mighty-Bo--sstones",'-')
     'Mighty-Mighty-Bo-sstones'
     """
 
     if str_.find(replace * 2) == -1:
         return str_
     else:
         str_ = str_.replace(replace * 2, replace)
         return recursive_replace(str_, replace)
 
 
 def repo_name_slug(value):
     """
     Return slug of name of repository
     This function is called on each creation/modification
     of repository to prevent bad names in repo
     """
 
     slug = remove_formatting(value)
     slug = strip_tags(slug)
 
     for c in """`?=[]\;'"<>,/~!@#$%^&*()+{}|: """:
         slug = slug.replace(c, '-')
     slug = recursive_replace(slug, '-')
     slug = collapse(slug, '-')
     return slug
 
 
 def get_repo_slug(request):
     _repo = request.environ['pylons.routes_dict'].get('repo_name')
     if _repo:
         _repo = _repo.rstrip('/')
     return _repo
 
 
 def get_repos_group_slug(request):
     _group = request.environ['pylons.routes_dict'].get('group_name')
     if _group:
         _group = _group.rstrip('/')
     return _group
 
 
 def action_logger(user, action, repo, ipaddr='', sa=None, commit=False):
     """
     Action logger for various actions made by users
 
     :param user: user that made this action, can be a unique username string or
         object containing user_id attribute
     :param action: action to log, should be on of predefined unique actions for
         easy translations
     :param repo: string name of repository or object containing repo_id,
         that action was made on
     :param ipaddr: optional ip address from what the action was made
     :param sa: optional sqlalchemy session
 
     """
 
     if not sa:
         sa = meta.Session()
 
     try:
         if hasattr(user, 'user_id'):
             user_obj = User.get(user.user_id)
         elif isinstance(user, basestring):
             user_obj = User.get_by_username(user)
         else:
             raise Exception('You have to provide a user object or a username')
 
         if hasattr(repo, 'repo_id'):
             repo_obj = Repository.get(repo.repo_id)
             repo_name = repo_obj.repo_name
         elif  isinstance(repo, basestring):
             repo_name = repo.lstrip('/')
             repo_obj = Repository.get_by_repo_name(repo_name)
         else:
             repo_obj = None
             repo_name = ''
 
         user_log = UserLog()
         user_log.user_id = user_obj.user_id
         user_log.username = user_obj.username
         user_log.action = safe_unicode(action)
 
         user_log.repository = repo_obj
         user_log.repository_name = repo_name
 
         user_log.action_date = datetime.datetime.now()
         user_log.user_ip = ipaddr
         sa.add(user_log)
 
         log.info('Logging action %s on %s by %s' %
                  (action, safe_unicode(repo), user_obj))
         if commit:
             sa.commit()
     except:
         log.error(traceback.format_exc())
         raise
 
 
 def get_repos(path, recursive=False, skip_removed_repos=True):
     """
     Scans given path for repos and return (name,(type,path)) tuple
 
     :param path: path to scan for repositories
     :param recursive: recursive search and return names with subdirs in front
     """
 
     # remove ending slash for better results
     path = path.rstrip(os.sep)
     log.debug('now scanning in %s location recursive:%s...' % (path, recursive))
 
     def _get_repos(p):
         if not os.access(p, os.W_OK):
             return
         for dirpath in os.listdir(p):
             if os.path.isfile(os.path.join(p, dirpath)):
                 continue
             cur_path = os.path.join(p, dirpath)
 
             # skip removed repos
             if skip_removed_repos and REMOVED_REPO_PAT.match(dirpath):
                 continue
 
             #skip .<somethin> dirs
             if dirpath.startswith('.'):
                 continue
 
             try:
                 scm_info = get_scm(cur_path)
                 yield scm_info[1].split(path, 1)[-1].lstrip(os.sep), scm_info
             except VCSError:
                 if not recursive:
                     continue
                 #check if this dir containts other repos for recursive scan
                 rec_path = os.path.join(p, dirpath)
                 if os.path.isdir(rec_path):
                     for inner_scm in _get_repos(rec_path):
                         yield inner_scm
 
     return _get_repos(path)
 
 #alias for backward compat
 get_filesystem_repos = get_repos
 
 
 def is_valid_repo(repo_name, base_path, scm=None):
     """
     Returns True if given path is a valid repository False otherwise.
     If scm param is given also compare if given scm is the same as expected
     from scm parameter
 
     :param repo_name:
     :param base_path:
     :param scm:
 
     :return True: if given path is a valid repository
     """
     full_path = os.path.join(safe_str(base_path), safe_str(repo_name))
 
     try:
         scm_ = get_scm(full_path)
         if scm:
             return scm_[0] == scm
         return True
     except VCSError:
         return False
 
 
 def is_valid_repos_group(repos_group_name, base_path):
     """
     Returns True if given path is a repos group False otherwise
 
     :param repo_name:
     :param base_path:
     """
     full_path = os.path.join(safe_str(base_path), safe_str(repos_group_name))
 
     # check if it's not a repo
     if is_valid_repo(repos_group_name, base_path):
         return False
 
     try:
         # we need to check bare git repos at higher level
         # since we might match branches/hooks/info/objects or possible
         # other things inside bare git repo
         get_scm(os.path.dirname(full_path))
         return False
     except VCSError:
         pass
 
     # check if it's a valid path
     if os.path.isdir(full_path):
         return True
 
     return False
 
 
 def ask_ok(prompt, retries=4, complaint='Yes or no please!'):
     while True:
         ok = raw_input(prompt)
         if ok in ('y', 'ye', 'yes'):
             return True
         if ok in ('n', 'no', 'nop', 'nope'):
             return False
         retries = retries - 1
         if retries < 0:
             raise IOError
         print complaint
 
 #propagated from mercurial documentation
 ui_sections = ['alias', 'auth',
                 'decode/encode', 'defaults',
                 'diff', 'email',
                 'extensions', 'format',
                 'merge-patterns', 'merge-tools',
                 'hooks', 'http_proxy',
                 'smtp', 'patch',
                 'paths', 'profiling',
                 'server', 'trusted',
                 'ui', 'web', ]
 
 
 def make_ui(read_from='file', path=None, checkpaths=True, clear_session=True):
     """
     A function that will read python rc files or database
     and make an mercurial ui object from read options
 
     :param path: path to mercurial config file
     :param checkpaths: check the path
     :param read_from: read from 'file' or 'db'
     """
 
     baseui = ui.ui()
 
     # clean the baseui object
     baseui._ocfg = config.config()
     baseui._ucfg = config.config()
     baseui._tcfg = config.config()
 
     if read_from == 'file':
         if not os.path.isfile(path):
             log.debug('hgrc file is not present at %s, skipping...' % path)
             return False
         log.debug('reading hgrc from %s' % path)
         cfg = config.config()
         cfg.read(path)
         for section in ui_sections:
             for k, v in cfg.items(section):
                 log.debug('settings ui from file: [%s] %s=%s' % (section, k, v))
                 baseui.setconfig(safe_str(section), safe_str(k), safe_str(v))
 
     elif read_from == 'db':
         sa = meta.Session()
         ret = sa.query(RhodeCodeUi)\
             .options(FromCache("sql_cache_short", "get_hg_ui_settings"))\
             .all()
 
         hg_ui = ret
         for ui_ in hg_ui:
             if ui_.ui_active:
                 log.debug('settings ui from db: [%s] %s=%s', ui_.ui_section,
                           ui_.ui_key, ui_.ui_value)
                 baseui.setconfig(safe_str(ui_.ui_section), safe_str(ui_.ui_key),
                                  safe_str(ui_.ui_value))
             if ui_.ui_key == 'push_ssl':
                 # force set push_ssl requirement to False, rhodecode
                 # handles that
                 baseui.setconfig(safe_str(ui_.ui_section), safe_str(ui_.ui_key),
                                  False)
         if clear_session:
             meta.Session.remove()
     return baseui
 
 
 def set_rhodecode_config(config):
     """
     Updates pylons config with new settings from database
 
     :param config:
     """
     hgsettings = RhodeCodeSetting.get_app_settings()
 
     for k, v in hgsettings.items():
         config[k] = v
 
 
 def invalidate_cache(cache_key, *args):
     """
     Puts cache invalidation task into db for
     further global cache invalidation
     """
 
     from rhodecode.model.scm import ScmModel
 
     if cache_key.startswith('get_repo_cached_'):
         name = cache_key.split('get_repo_cached_')[-1]
         ScmModel().mark_for_invalidation(name)
 
 
 def map_groups(path):
     """
     Given a full path to a repository, create all nested groups that this
     repo is inside. This function creates parent-child relationships between
     groups and creates default perms for all new groups.
 
     :param paths: full path to repository
     """
     sa = meta.Session()
     groups = path.split(Repository.url_sep())
     parent = None
     group = None
 
     # last element is repo in nested groups structure
     groups = groups[:-1]
     rgm = ReposGroupModel(sa)
     for lvl, group_name in enumerate(groups):
         group_name = '/'.join(groups[:lvl] + [group_name])
         group = RepoGroup.get_by_group_name(group_name)
         desc = '%s group' % group_name
 
         # skip folders that are now removed repos
         if REMOVED_REPO_PAT.match(group_name):
             break
 
         if group is None:
             log.debug('creating group level: %s group_name: %s' % (lvl,
                                                                    group_name))
             group = RepoGroup(group_name, parent)
             group.group_description = desc
             sa.add(group)
             rgm._create_default_perms(group)
             sa.flush()
         parent = group
     return group
 
 
 def repo2db_mapper(initial_repo_list, remove_obsolete=False,
                    install_git_hook=False):
     """
     maps all repos given in initial_repo_list, non existing repositories
     are created, if remove_obsolete is True it also check for db entries
     that are not in initial_repo_list and removes them.
 
     :param initial_repo_list: list of repositories found by scanning methods
     :param remove_obsolete: check for obsolete entries in database
     :param install_git_hook: if this is True, also check and install githook
         for a repo if missing
     """
     from rhodecode.model.repo import RepoModel
     from rhodecode.model.scm import ScmModel
     sa = meta.Session()
     rm = RepoModel()
     user = sa.query(User).filter(User.admin == True).first()
     if user is None:
         raise Exception('Missing administrative account!')
     added = []
 
 #    # clear cache keys
 #    log.debug("Clearing cache keys now...")
 #    CacheInvalidation.clear_cache()
 #    sa.commit()
 
     ##creation defaults
     defs = RhodeCodeSetting.get_default_repo_settings(strip_prefix=True)
     enable_statistics = defs.get('repo_enable_statistics')
     enable_locking = defs.get('repo_enable_locking')
     enable_downloads = defs.get('repo_enable_downloads')
     private = defs.get('repo_private')
 
     for name, repo in initial_repo_list.items():
         group = map_groups(name)
         db_repo = rm.get_by_repo_name(name)
         # found repo that is on filesystem not in RhodeCode database
         if not db_repo:
             log.info('repository %s not found, creating now' % name)
             added.append(name)
             desc = (repo.description
                     if repo.description != 'unknown'
                     else '%s repository' % name)
 
             new_repo = rm.create_repo(
                 repo_name=name,
                 repo_type=repo.alias,
                 description=desc,
                 repos_group=getattr(group, 'group_id', None),
                 owner=user,
                 just_db=True,
                 enable_locking=enable_locking,
                 enable_downloads=enable_downloads,
                 enable_statistics=enable_statistics,
                 private=private
             )
             # we added that repo just now, and make sure it has githook
             # installed
             if new_repo.repo_type == 'git':
                 ScmModel().install_git_hook(new_repo.scm_instance)
             new_repo.update_changeset_cache()
         elif install_git_hook:
             if db_repo.repo_type == 'git':
                 ScmModel().install_git_hook(db_repo.scm_instance)
         # during starting install all cache keys for all repositories in the
         # system, this will register all repos and multiple instances
         key, _prefix, _org_key = CacheInvalidation._get_key(name)
         CacheInvalidation.invalidate(name)
         log.debug("Creating a cache key for %s, instance_id %s"
                   % (name, _prefix or 'unknown'))
 
     sa.commit()
     removed = []
     if remove_obsolete:
         # remove from database those repositories that are not in the filesystem
         for repo in sa.query(Repository).all():
             if repo.repo_name not in initial_repo_list.keys():
                 log.debug("Removing non-existing repository found in db `%s`" %
                           repo.repo_name)
                 try:
                     sa.delete(repo)
                     sa.commit()
                     removed.append(repo.repo_name)
                 except:
                     #don't hold further removals on error
                     log.error(traceback.format_exc())
                     sa.rollback()
 
     return added, removed
 
 
 # set cache regions for beaker so celery can utilise it
 def add_cache(settings):
     cache_settings = {'regions': None}
     for key in settings.keys():
         for prefix in ['beaker.cache.', 'cache.']:
             if key.startswith(prefix):
                 name = key.split(prefix)[1].strip()
                 cache_settings[name] = settings[key].strip()
     if cache_settings['regions']:
         for region in cache_settings['regions'].split(','):
             region = region.strip()
             region_settings = {}
             for key, value in cache_settings.items():
                 if key.startswith(region):
                     region_settings[key.split('.')[1]] = value
             region_settings['expire'] = int(region_settings.get('expire',
                                                                 60))
             region_settings.setdefault('lock_dir',
                                        cache_settings.get('lock_dir'))
             region_settings.setdefault('data_dir',
                                        cache_settings.get('data_dir'))
 
             if 'type' not in region_settings:
                 region_settings['type'] = cache_settings.get('type',
                                                              'memory')
             beaker.cache.cache_regions[region] = region_settings
 
 
 def load_rcextensions(root_path):
     import rhodecode
     from rhodecode.config import conf
 
     path = os.path.join(root_path, 'rcextensions', '__init__.py')
     if os.path.isfile(path):
         rcext = create_module('rc', path)
         EXT = rhodecode.EXTENSIONS = rcext
         log.debug('Found rcextensions now loading %s...' % rcext)
 
         # Additional mappings that are not present in the pygments lexers
         conf.LANGUAGES_EXTENSIONS_MAP.update(getattr(EXT, 'EXTRA_MAPPINGS', {}))
 
         #OVERRIDE OUR EXTENSIONS FROM RC-EXTENSIONS (if present)
 
         if getattr(EXT, 'INDEX_EXTENSIONS', []) != []:
             log.debug('settings custom INDEX_EXTENSIONS')
             conf.INDEX_EXTENSIONS = getattr(EXT, 'INDEX_EXTENSIONS', [])
 
         #ADDITIONAL MAPPINGS
         log.debug('adding extra into INDEX_EXTENSIONS')
         conf.INDEX_EXTENSIONS.extend(getattr(EXT, 'EXTRA_INDEX_EXTENSIONS', []))
 
         # auto check if the module is not missing any data, set to default if is
         # this will help autoupdate new feature of rcext module
         from rhodecode.config import rcextensions
         for k in dir(rcextensions):
             if not k.startswith('_') and not hasattr(EXT, k):
                 setattr(EXT, k, getattr(rcextensions, k))
 
 
 def get_custom_lexer(extension):
     """
     returns a custom lexer if it's defined in rcextensions module, or None
     if there's no custom lexer defined
     """
     import rhodecode
     from pygments import lexers
     #check if we didn't define this extension as other lexer
     if rhodecode.EXTENSIONS and extension in rhodecode.EXTENSIONS.EXTRA_LEXERS:
         _lexer_name = rhodecode.EXTENSIONS.EXTRA_LEXERS[extension]
         return lexers.get_lexer_by_name(_lexer_name)
 
 
 #==============================================================================
 # TEST FUNCTIONS AND CREATORS
 #==============================================================================
 def create_test_index(repo_location, config, full_index):
     """
     Makes default test index
 
     :param config: test config
     :param full_index:
     """
 
     from rhodecode.lib.indexers.daemon import WhooshIndexingDaemon
     from rhodecode.lib.pidlock import DaemonLock, LockHeld
 
     repo_location = repo_location
 
     index_location = os.path.join(config['app_conf']['index_dir'])
     if not os.path.exists(index_location):
         os.makedirs(index_location)
 
     try:
         l = DaemonLock(file_=jn(dn(index_location), 'make_index.lock'))
         WhooshIndexingDaemon(index_location=index_location,
                              repo_location=repo_location)\
             .run(full_index=full_index)
         l.release()
     except LockHeld:
         pass
 
 
 def create_test_env(repos_test_path, config):
     """
     Makes a fresh database and
     install test repository into tmp dir
     """
     from rhodecode.lib.db_manage import DbManage
     from rhodecode.tests import HG_REPO, GIT_REPO, TESTS_TMP_PATH
 
     # PART ONE create db
     dbconf = config['sqlalchemy.db1.url']
     log.debug('making test db %s' % dbconf)
 
     # create test dir if it doesn't exist
     if not os.path.isdir(repos_test_path):
         log.debug('Creating testdir %s' % repos_test_path)
         os.makedirs(repos_test_path)
 
     dbmanage = DbManage(log_sql=True, dbconf=dbconf, root=config['here'],
                         tests=True)
     dbmanage.create_tables(override=True)
     dbmanage.create_settings(dbmanage.config_prompt(repos_test_path))
     dbmanage.create_default_user()
     dbmanage.admin_prompt()
     dbmanage.create_permissions()
     dbmanage.populate_default_permissions()
     Session().commit()
     # PART TWO make test repo
     log.debug('making test vcs repositories')
 
     idx_path = config['app_conf']['index_dir']
     data_path = config['app_conf']['cache_dir']
 
     #clean index and data
     if idx_path and os.path.exists(idx_path):
         log.debug('remove %s' % idx_path)
         shutil.rmtree(idx_path)
 
     if data_path and os.path.exists(data_path):
         log.debug('remove %s' % data_path)
         shutil.rmtree(data_path)
 
     #CREATE DEFAULT TEST REPOS
     cur_dir = dn(dn(abspath(__file__)))
     tar = tarfile.open(jn(cur_dir, 'tests', "vcs_test_hg.tar.gz"))
     tar.extractall(jn(TESTS_TMP_PATH, HG_REPO))
     tar.close()
 
     cur_dir = dn(dn(abspath(__file__)))
     tar = tarfile.open(jn(cur_dir, 'tests', "vcs_test_git.tar.gz"))
     tar.extractall(jn(TESTS_TMP_PATH, GIT_REPO))
     tar.close()
 
     #LOAD VCS test stuff
     from rhodecode.tests.vcs import setup_package
     setup_package()
 
 
 #==============================================================================
 # PASTER COMMANDS
 #==============================================================================
 class BasePasterCommand(Command):
     """
     Abstract Base Class for paster commands.
 
     The celery commands are somewhat aggressive about loading
     celery.conf, and since our module sets the `CELERY_LOADER`
     environment variable to our loader, we have to bootstrap a bit and
     make sure we've had a chance to load the pylons config off of the
     command line, otherwise everything fails.
     """
     min_args = 1
     min_args_error = "Please provide a paster config file as an argument."
     takes_config_file = 1
     requires_config_file = True
 
     def notify_msg(self, msg, log=False):
         """Make a notification to user, additionally if logger is passed
         it logs this action using given logger
 
         :param msg: message that will be printed to user
         :param log: logging instance, to use to additionally log this message
 
         """
         if log and isinstance(log, logging):
             log(msg)
 
     def run(self, args):
         """
         Overrides Command.run
 
         Checks for a config file argument and loads it.
         """
         if len(args) < self.min_args:
             raise BadCommand(
                 self.min_args_error % {'min_args': self.min_args,
                                        'actual_args': len(args)})
 
         # Decrement because we're going to lob off the first argument.
         # @@ This is hacky
         self.min_args -= 1
         self.bootstrap_config(args[0])
         self.update_parser()
         return super(BasePasterCommand, self).run(args[1:])
 
     def update_parser(self):
         """
         Abstract method.  Allows for the class's parser to be updated
         before the superclass's `run` method is called.  Necessary to
         allow options/arguments to be passed through to the underlying
         celery command.
         """
         raise NotImplementedError("Abstract Method.")
 
     def bootstrap_config(self, conf):
         """
         Loads the pylons configuration.
         """
         from pylons import config as pylonsconfig
 
         self.path_to_ini_file = os.path.realpath(conf)
         conf = paste.deploy.appconfig('config:' + self.path_to_ini_file)
         pylonsconfig.init_app(conf.global_conf, conf.local_conf)
 
     def _init_session(self):
         """
         Inits SqlAlchemy Session
         """
         logging.config.fileConfig(self.path_to_ini_file)
         from pylons import config
         from rhodecode.model import init_model
         from rhodecode.lib.utils2 import engine_from_config
 
         #get to remove repos !!
         add_cache(config)
         engine = engine_from_config(config, 'sqlalchemy.db1.')
         init_model(engine)
 
 
 def check_git_version():
     """
     Checks what version of git is installed in system, and issues a warning
     if it's too old for RhodeCode to properly work.
     """
-    import subprocess
+    from rhodecode import BACKENDS
+    from rhodecode.lib.vcs.backends.git.repository import GitRepository
     from distutils.version import StrictVersion
-    from rhodecode import BACKENDS
 
-    p = subprocess.Popen('git --version', shell=True,
-                        stdout=subprocess.PIPE, stderr=subprocess.PIPE)
-    stdout, stderr = p.communicate()
+    stdout, stderr = GitRepository._run_git_command('--version')
+
     ver = (stdout.split(' ')[-1] or '').strip() or '0.0.0'
     if len(ver.split('.')) > 3:
         #StrictVersion needs to be only 3 element type
         ver = '.'.join(ver.split('.')[:3])
     try:
         _ver = StrictVersion(ver)
     except:
         _ver = StrictVersion('0.0.0')
         stderr = traceback.format_exc()
 
     req_ver = '1.7.4'
     to_old_git = False
     if  _ver < StrictVersion(req_ver):
         to_old_git = True
 
     if 'git' in BACKENDS:
         log.debug('GIT version detected: %s' % stdout)
         if stderr:
             log.warning('Unable to detect git version org error was:%r' % stderr)
         elif to_old_git:
             log.warning('RhodeCode detected git version %s, which is too old '
                         'for the system to function properly. Make sure '
                         'its version is at least %s' % (ver, req_ver))
     return _ver
 
 
 @decorator.decorator
 def jsonify(func, *args, **kwargs):
     """Action decorator that formats output for JSON
 
     Given a function that will return content, this decorator will turn
     the result into JSON, with a content-type of 'application/json' and
     output it.
 
     """
     from pylons.decorators.util import get_pylons
     from rhodecode.lib.ext_json import json
     pylons = get_pylons(args)
     pylons.response.headers['Content-Type'] = 'application/json; charset=utf-8'
     data = func(*args, **kwargs)
     if isinstance(data, (list, tuple)):
         msg = "JSON responses with Array envelopes are susceptible to " \
               "cross-site data leak attacks, see " \
               "http://wiki.pylonshq.com/display/pylonsfaq/Warnings"
         warnings.warn(msg, Warning, 2)
         log.warning(msg)
     log.debug("Returning JSON wrapped action output")
     return json.dumps(data, encoding='utf-8')

rhodecode/lib/vcs/backends/git/changeset.py

@@ -1,542 +1,544 b''
 import re
 from itertools import chain
 from dulwich import objects
 from subprocess import Popen, PIPE
+import rhodecode
 from rhodecode.lib.vcs.conf import settings
 from rhodecode.lib.vcs.exceptions import RepositoryError
 from rhodecode.lib.vcs.exceptions import ChangesetError
 from rhodecode.lib.vcs.exceptions import NodeDoesNotExistError
 from rhodecode.lib.vcs.exceptions import VCSError
 from rhodecode.lib.vcs.exceptions import ChangesetDoesNotExistError
 from rhodecode.lib.vcs.exceptions import ImproperArchiveTypeError
 from rhodecode.lib.vcs.backends.base import BaseChangeset, EmptyChangeset
 from rhodecode.lib.vcs.nodes import FileNode, DirNode, NodeKind, RootNode, \
     RemovedFileNode, SubModuleNode, ChangedFileNodesGenerator,\
     AddedFileNodesGenerator, RemovedFileNodesGenerator
 from rhodecode.lib.vcs.utils import safe_unicode
 from rhodecode.lib.vcs.utils import date_fromtimestamp
 from rhodecode.lib.vcs.utils.lazy import LazyProperty
 
 
 class GitChangeset(BaseChangeset):
     """
     Represents state of the repository at single revision.
     """
 
     def __init__(self, repository, revision):
         self._stat_modes = {}
         self.repository = repository
 
         try:
             commit = self.repository._repo.get_object(revision)
             if isinstance(commit, objects.Tag):
                 revision = commit.object[1]
                 commit = self.repository._repo.get_object(commit.object[1])
         except KeyError:
             raise RepositoryError("Cannot get object with id %s" % revision)
         self.raw_id = revision
         self.id = self.raw_id
         self.short_id = self.raw_id[:12]
         self._commit = commit
 
         self._tree_id = commit.tree
         self._commiter_property = 'committer'
         self._author_property = 'author'
         self._date_property = 'commit_time'
         self._date_tz_property = 'commit_timezone'
         self.revision = repository.revisions.index(revision)
 
         self.message = safe_unicode(commit.message)
 
         self.nodes = {}
         self._paths = {}
 
     @LazyProperty
     def commiter(self):
         return safe_unicode(getattr(self._commit, self._commiter_property))
 
     @LazyProperty
     def author(self):
         return safe_unicode(getattr(self._commit, self._author_property))
 
     @LazyProperty
     def date(self):
         return date_fromtimestamp(getattr(self._commit, self._date_property),
                                   getattr(self._commit, self._date_tz_property))
 
     @LazyProperty
     def _timestamp(self):
         return getattr(self._commit, self._date_property)
 
     @LazyProperty
     def status(self):
         """
         Returns modified, added, removed, deleted files for current changeset
         """
         return self.changed, self.added, self.removed
 
     @LazyProperty
     def tags(self):
         _tags = []
         for tname, tsha in self.repository.tags.iteritems():
             if tsha == self.raw_id:
                 _tags.append(tname)
         return _tags
 
     @LazyProperty
     def branch(self):
 
         heads = self.repository._heads(reverse=False)
 
         ref = heads.get(self.raw_id)
         if ref:
             return safe_unicode(ref)
 
     def _fix_path(self, path):
         """
         Paths are stored without trailing slash so we need to get rid off it if
         needed.
         """
         if path.endswith('/'):
             path = path.rstrip('/')
         return path
 
     def _get_id_for_path(self, path):
 
         # FIXME: Please, spare a couple of minutes and make those codes cleaner;
         if not path in self._paths:
             path = path.strip('/')
             # set root tree
             tree = self.repository._repo[self._tree_id]
             if path == '':
                 self._paths[''] = tree.id
                 return tree.id
             splitted = path.split('/')
             dirs, name = splitted[:-1], splitted[-1]
             curdir = ''
 
             # initially extract things from root dir
             for item, stat, id in tree.iteritems():
                 if curdir:
                     name = '/'.join((curdir, item))
                 else:
                     name = item
                 self._paths[name] = id
                 self._stat_modes[name] = stat
 
             for dir in dirs:
                 if curdir:
                     curdir = '/'.join((curdir, dir))
                 else:
                     curdir = dir
                 dir_id = None
                 for item, stat, id in tree.iteritems():
                     if dir == item:
                         dir_id = id
                 if dir_id:
                     # Update tree
                     tree = self.repository._repo[dir_id]
                     if not isinstance(tree, objects.Tree):
                         raise ChangesetError('%s is not a directory' % curdir)
                 else:
                     raise ChangesetError('%s have not been found' % curdir)
 
                 # cache all items from the given traversed tree
                 for item, stat, id in tree.iteritems():
                     if curdir:
                         name = '/'.join((curdir, item))
                     else:
                         name = item
                     self._paths[name] = id
                     self._stat_modes[name] = stat
             if not path in self._paths:
                 raise NodeDoesNotExistError("There is no file nor directory "
                     "at the given path %r at revision %r"
                     % (path, self.short_id))
         return self._paths[path]
 
     def _get_kind(self, path):
         obj = self.repository._repo[self._get_id_for_path(path)]
         if isinstance(obj, objects.Blob):
             return NodeKind.FILE
         elif isinstance(obj, objects.Tree):
             return NodeKind.DIR
 
     def _get_filectx(self, path):
         path = self._fix_path(path)
         if self._get_kind(path) != NodeKind.FILE:
             raise ChangesetError("File does not exist for revision %r at "
                 " %r" % (self.raw_id, path))
         return path
 
     def _get_file_nodes(self):
         return chain(*(t[2] for t in self.walk()))
 
     @LazyProperty
     def parents(self):
         """
         Returns list of parents changesets.
         """
         return [self.repository.get_changeset(parent)
                 for parent in self._commit.parents]
 
     @LazyProperty
     def children(self):
         """
         Returns list of children changesets.
         """
         so, se = self.repository.run_git_command(
             "rev-list --all --children | grep '^%s'" % self.raw_id
         )
 
         children = []
         for l in so.splitlines():
             childs = l.split(' ')[1:]
             children.extend(childs)
         return [self.repository.get_changeset(cs) for cs in children]
 
     def next(self, branch=None):
 
         if branch and self.branch != branch:
             raise VCSError('Branch option used on changeset not belonging '
                            'to that branch')
 
         def _next(changeset, branch):
             try:
                 next_ = changeset.revision + 1
                 next_rev = changeset.repository.revisions[next_]
             except IndexError:
                 raise ChangesetDoesNotExistError
             cs = changeset.repository.get_changeset(next_rev)
 
             if branch and branch != cs.branch:
                 return _next(cs, branch)
 
             return cs
 
         return _next(self, branch)
 
     def prev(self, branch=None):
         if branch and self.branch != branch:
             raise VCSError('Branch option used on changeset not belonging '
                            'to that branch')
 
         def _prev(changeset, branch):
             try:
                 prev_ = changeset.revision - 1
                 if prev_ < 0:
                     raise IndexError
                 prev_rev = changeset.repository.revisions[prev_]
             except IndexError:
                 raise ChangesetDoesNotExistError
 
             cs = changeset.repository.get_changeset(prev_rev)
 
             if branch and branch != cs.branch:
                 return _prev(cs, branch)
 
             return cs
 
         return _prev(self, branch)
 
     def diff(self, ignore_whitespace=True, context=3):
         rev1 = self.parents[0] if self.parents else self.repository.EMPTY_CHANGESET
         rev2 = self
         return ''.join(self.repository.get_diff(rev1, rev2,
                                     ignore_whitespace=ignore_whitespace,
                                     context=context))
 
     def get_file_mode(self, path):
         """
         Returns stat mode of the file at the given ``path``.
         """
         # ensure path is traversed
         self._get_id_for_path(path)
         return self._stat_modes[path]
 
     def get_file_content(self, path):
         """
         Returns content of the file at given ``path``.
         """
         id = self._get_id_for_path(path)
         blob = self.repository._repo[id]
         return blob.as_pretty_string()
 
     def get_file_size(self, path):
         """
         Returns size of the file at given ``path``.
         """
         id = self._get_id_for_path(path)
         blob = self.repository._repo[id]
         return blob.raw_length()
 
     def get_file_changeset(self, path):
         """
         Returns last commit of the file at the given ``path``.
         """
         node = self.get_node(path)
         return node.history[0]
 
     def get_file_history(self, path):
         """
         Returns history of file as reversed list of ``Changeset`` objects for
         which file at given ``path`` has been modified.
 
         TODO: This function now uses os underlying 'git' and 'grep' commands
         which is generally not good. Should be replaced with algorithm
         iterating commits.
         """
         self._get_filectx(path)
 
         cmd = 'log --pretty="format: %%H" -s -p %s -- "%s"' % (
                   self.id, path
                )
         so, se = self.repository.run_git_command(cmd)
         ids = re.findall(r'[0-9a-fA-F]{40}', so)
         return [self.repository.get_changeset(id) for id in ids]
 
     def get_file_history_2(self, path):
         """
         Returns history of file as reversed list of ``Changeset`` objects for
         which file at given ``path`` has been modified.
 
         """
         self._get_filectx(path)
         from dulwich.walk import Walker
         include = [self.id]
         walker = Walker(self.repository._repo.object_store, include,
                         paths=[path], max_entries=1)
         return [self.repository.get_changeset(sha)
                 for sha in (x.commit.id for x in walker)]
 
     def get_file_annotate(self, path):
         """
         Returns a generator of four element tuples with
             lineno, sha, changeset lazy loader and line
 
         TODO: This function now uses os underlying 'git' command which is
         generally not good. Should be replaced with algorithm iterating
         commits.
         """
         cmd = 'blame -l --root -r %s -- "%s"' % (self.id, path)
         # -l     ==> outputs long shas (and we need all 40 characters)
         # --root ==> doesn't put '^' character for bounderies
         # -r sha ==> blames for the given revision
         so, se = self.repository.run_git_command(cmd)
 
         for i, blame_line in enumerate(so.split('\n')[:-1]):
             ln_no = i + 1
             sha, line = re.split(r' ', blame_line, 1)
             yield (ln_no, sha, lambda: self.repository.get_changeset(sha), line)
 
     def fill_archive(self, stream=None, kind='tgz', prefix=None,
                      subrepos=False):
         """
         Fills up given stream.
 
         :param stream: file like object.
         :param kind: one of following: ``zip``, ``tgz`` or ``tbz2``.
             Default: ``tgz``.
         :param prefix: name of root directory in archive.
             Default is repository name and changeset's raw_id joined with dash
             (``repo-tip.<KIND>``).
         :param subrepos: include subrepos in this archive.
 
         :raise ImproperArchiveTypeError: If given kind is wrong.
         :raise VcsError: If given stream is None
 
         """
         allowed_kinds = settings.ARCHIVE_SPECS.keys()
         if kind not in allowed_kinds:
             raise ImproperArchiveTypeError('Archive kind not supported use one'
                 'of %s', allowed_kinds)
 
         if prefix is None:
             prefix = '%s-%s' % (self.repository.name, self.short_id)
         elif prefix.startswith('/'):
             raise VCSError("Prefix cannot start with leading slash")
         elif prefix.strip() == '':
             raise VCSError("Prefix cannot be empty")
 
         if kind == 'zip':
             frmt = 'zip'
         else:
             frmt = 'tar'
-        cmd = 'git archive --format=%s --prefix=%s/ %s' % (frmt, prefix,
-            self.raw_id)
+        _git_path = rhodecode.CONFIG.get('git_path', 'git')
+        cmd = '%s archive --format=%s --prefix=%s/ %s' % (_git_path, 
+                                                frmt, prefix, self.raw_id)
         if kind == 'tgz':
             cmd += ' | gzip -9'
         elif kind == 'tbz2':
             cmd += ' | bzip2 -9'
 
         if stream is None:
             raise VCSError('You need to pass in a valid stream for filling'
                            ' with archival data')
         popen = Popen(cmd, stdout=PIPE, stderr=PIPE, shell=True,
             cwd=self.repository.path)
 
         buffer_size = 1024 * 8
         chunk = popen.stdout.read(buffer_size)
         while chunk:
             stream.write(chunk)
             chunk = popen.stdout.read(buffer_size)
         # Make sure all descriptors would be read
         popen.communicate()
 
     def get_nodes(self, path):
         if self._get_kind(path) != NodeKind.DIR:
             raise ChangesetError("Directory does not exist for revision %r at "
                 " %r" % (self.revision, path))
         path = self._fix_path(path)
         id = self._get_id_for_path(path)
         tree = self.repository._repo[id]
         dirnodes = []
         filenodes = []
         als = self.repository.alias
         for name, stat, id in tree.iteritems():
             if objects.S_ISGITLINK(stat):
                 dirnodes.append(SubModuleNode(name, url=None, changeset=id,
                                               alias=als))
                 continue
 
             obj = self.repository._repo.get_object(id)
             if path != '':
                 obj_path = '/'.join((path, name))
             else:
                 obj_path = name
             if obj_path not in self._stat_modes:
                 self._stat_modes[obj_path] = stat
             if isinstance(obj, objects.Tree):
                 dirnodes.append(DirNode(obj_path, changeset=self))
             elif isinstance(obj, objects.Blob):
                 filenodes.append(FileNode(obj_path, changeset=self, mode=stat))
             else:
                 raise ChangesetError("Requested object should be Tree "
                                      "or Blob, is %r" % type(obj))
         nodes = dirnodes + filenodes
         for node in nodes:
             if not node.path in self.nodes:
                 self.nodes[node.path] = node
         nodes.sort()
         return nodes
 
     def get_node(self, path):
         if isinstance(path, unicode):
             path = path.encode('utf-8')
         path = self._fix_path(path)
         if not path in self.nodes:
             try:
                 id_ = self._get_id_for_path(path)
             except ChangesetError:
                 raise NodeDoesNotExistError("Cannot find one of parents' "
                     "directories for a given path: %s" % path)
 
             _GL = lambda m: m and objects.S_ISGITLINK(m)
             if _GL(self._stat_modes.get(path)):
                 node = SubModuleNode(path, url=None, changeset=id_,
                                      alias=self.repository.alias)
             else:
                 obj = self.repository._repo.get_object(id_)
 
                 if isinstance(obj, objects.Tree):
                     if path == '':
                         node = RootNode(changeset=self)
                     else:
                         node = DirNode(path, changeset=self)
                     node._tree = obj
                 elif isinstance(obj, objects.Blob):
                     node = FileNode(path, changeset=self)
                     node._blob = obj
                 else:
                     raise NodeDoesNotExistError("There is no file nor directory "
                         "at the given path %r at revision %r"
                         % (path, self.short_id))
             # cache node
             self.nodes[path] = node
         return self.nodes[path]
 
     @LazyProperty
     def affected_files(self):
         """
         Get's a fast accessible file changes for given changeset
         """
         a, m, d = self._changes_cache
         return list(a.union(m).union(d))
 
     @LazyProperty
     def _diff_name_status(self):
         output = []
         for parent in self.parents:
             cmd = 'diff --name-status %s %s --encoding=utf8' % (parent.raw_id,
                                                                 self.raw_id)
             so, se = self.repository.run_git_command(cmd)
             output.append(so.strip())
         return '\n'.join(output)
 
     @LazyProperty
     def _changes_cache(self):
         added = set()
         modified = set()
         deleted = set()
         _r = self.repository._repo
 
         parents = self.parents
         if not self.parents:
             parents = [EmptyChangeset()]
         for parent in parents:
             if isinstance(parent, EmptyChangeset):
                 oid = None
             else:
                 oid = _r[parent.raw_id].tree
             changes = _r.object_store.tree_changes(oid, _r[self.raw_id].tree)
             for (oldpath, newpath), (_, _), (_, _) in changes:
                 if newpath and oldpath:
                     modified.add(newpath)
                 elif newpath and not oldpath:
                     added.add(newpath)
                 elif not newpath and oldpath:
                     deleted.add(oldpath)
         return added, modified, deleted
 
     def _get_paths_for_status(self, status):
         """
         Returns sorted list of paths for given ``status``.
 
         :param status: one of: *added*, *modified* or *deleted*
         """
         a, m, d = self._changes_cache
         return sorted({
             'added': list(a),
             'modified': list(m),
             'deleted': list(d)}[status]
         )
 
     @LazyProperty
     def added(self):
         """
         Returns list of added ``FileNode`` objects.
         """
         if not self.parents:
             return list(self._get_file_nodes())
         return AddedFileNodesGenerator([n for n in
                                 self._get_paths_for_status('added')], self)
 
     @LazyProperty
     def changed(self):
         """
         Returns list of modified ``FileNode`` objects.
         """
         if not self.parents:
             return []
         return ChangedFileNodesGenerator([n for n in
                                 self._get_paths_for_status('modified')], self)
 
     @LazyProperty
     def removed(self):
         """
         Returns list of removed ``FileNode`` objects.
         """
         if not self.parents:
             return []
         return RemovedFileNodesGenerator([n for n in
                                 self._get_paths_for_status('deleted')], self)

rhodecode/lib/vcs/backends/git/repository.py

@@ -1,669 +1,673 b''
 # -*- coding: utf-8 -*-
 """
     vcs.backends.git
     ~~~~~~~~~~~~~~~~
 
     Git backend implementation.
 
     :created_on: Apr 8, 2010
     :copyright: (c) 2010-2011 by Marcin Kuzminski, Lukasz Balcerzak.
 """
 
 import os
 import re
 import time
 import posixpath
 import logging
 import traceback
 import urllib
 import urllib2
 from dulwich.repo import Repo, NotGitRepository
 from dulwich.objects import Tag
 from string import Template
-from subprocess import Popen, PIPE
+
+import rhodecode
 from rhodecode.lib.vcs.backends.base import BaseRepository
 from rhodecode.lib.vcs.exceptions import BranchDoesNotExistError
 from rhodecode.lib.vcs.exceptions import ChangesetDoesNotExistError
 from rhodecode.lib.vcs.exceptions import EmptyRepositoryError
 from rhodecode.lib.vcs.exceptions import RepositoryError
 from rhodecode.lib.vcs.exceptions import TagAlreadyExistError
 from rhodecode.lib.vcs.exceptions import TagDoesNotExistError
 from rhodecode.lib.vcs.utils import safe_unicode, makedate, date_fromtimestamp
 from rhodecode.lib.vcs.utils.lazy import LazyProperty, ThreadLocalLazyProperty
 from rhodecode.lib.vcs.utils.ordered_dict import OrderedDict
 from rhodecode.lib.vcs.utils.paths import abspath
 from rhodecode.lib.vcs.utils.paths import get_user_home
 from .workdir import GitWorkdir
 from .changeset import GitChangeset
 from .inmemory import GitInMemoryChangeset
 from .config import ConfigFile
 from rhodecode.lib import subprocessio
 
 
 log = logging.getLogger(__name__)
 
 
 class GitRepository(BaseRepository):
     """
     Git repository backend.
     """
     DEFAULT_BRANCH_NAME = 'master'
     scm = 'git'
 
     def __init__(self, repo_path, create=False, src_url=None,
                  update_after_clone=False, bare=False):
 
         self.path = abspath(repo_path)
         repo = self._get_repo(create, src_url, update_after_clone, bare)
         self.bare = repo.bare
 
         self._config_files = [
             bare and abspath(self.path, 'config')
                      or abspath(self.path, '.git', 'config'),
             abspath(get_user_home(), '.gitconfig'),
         ]
 
     @ThreadLocalLazyProperty
     def _repo(self):
         repo = Repo(self.path)
         #temporary set that to now at later we will move it to constructor
         baseui = None
         if baseui is None:
             from mercurial.ui import ui
             baseui = ui()
         # patch the instance of GitRepo with an "FAKE" ui object to add
         # compatibility layer with Mercurial
         setattr(repo, 'ui', baseui)
         return repo
 
     @property
     def head(self):
         try:
             return self._repo.head()
         except KeyError:
             return None
 
     @LazyProperty
     def revisions(self):
         """
         Returns list of revisions' ids, in ascending order.  Being lazy
         attribute allows external tools to inject shas from cache.
         """
         return self._get_all_revisions()
 
-    def run_git_command(self, cmd):
+    @classmethod
+    def _run_git_command(cls, cmd, **opts):
         """
         Runs given ``cmd`` as git command and returns tuple
-        (returncode, stdout, stderr).
-
-        .. note::
-           This method exists only until log/blame functionality is implemented
-           at Dulwich (see https://bugs.launchpad.net/bugs/645142). Parsing
-           os command's output is road to hell...
+        (stdout, stderr).
 
         :param cmd: git command to be executed
+        :param opts: env options to pass into Subprocess command
         """
 
         _copts = ['-c', 'core.quotepath=false', ]
         _str_cmd = False
         if isinstance(cmd, basestring):
             cmd = [cmd]
             _str_cmd = True
 
         gitenv = os.environ
         # need to clean fix GIT_DIR !
         if 'GIT_DIR' in gitenv:
             del gitenv['GIT_DIR']
         gitenv['GIT_CONFIG_NOGLOBAL'] = '1'
 
-        cmd = ['git'] + _copts + cmd
+        _git_path = rhodecode.CONFIG.get('git_path', 'git')
+        cmd = [_git_path] + _copts + cmd
         if _str_cmd:
             cmd = ' '.join(cmd)
         try:
-            opts = dict(
+            _opts = dict(
                 env=gitenv,
                 shell=False,
             )
-            if os.path.isdir(self.path):
-                opts['cwd'] = self.path
-            p = subprocessio.SubprocessIOChunker(cmd, **opts)
+            _opts.update(opts)
+            p = subprocessio.SubprocessIOChunker(cmd, **_opts)
         except (EnvironmentError, OSError), err:
             log.error(traceback.format_exc())
             raise RepositoryError("Couldn't run git command (%s).\n"
                                   "Original error was:%s" % (cmd, err))
 
         return ''.join(p.output), ''.join(p.error)
 
+    def run_git_command(self, cmd):
+        opts = {}
+        if os.path.isdir(self.path):
+            opts['cwd'] = self.path
+        return self._run_git_command(cmd, **opts)
+
     @classmethod
     def _check_url(cls, url):
         """
         Functon will check given url and try to verify if it's a valid
         link. Sometimes it may happened that mercurial will issue basic
         auth request that can cause whole API to hang when used from python
         or other external calls.
 
         On failures it'll raise urllib2.HTTPError
         """
         from mercurial.util import url as Url
 
         # those authnadlers are patched for python 2.6.5 bug an
         # infinit looping when given invalid resources
         from mercurial.url import httpbasicauthhandler, httpdigestauthhandler
 
         # check first if it's not an local url
         if os.path.isdir(url) or url.startswith('file:'):
             return True
 
         if('+' in url[:url.find('://')]):
             url = url[url.find('+') + 1:]
 
         handlers = []
         test_uri, authinfo = Url(url).authinfo()
         if not test_uri.endswith('info/refs'):
             test_uri = test_uri.rstrip('/') + '/info/refs'
         if authinfo:
             #create a password manager
             passmgr = urllib2.HTTPPasswordMgrWithDefaultRealm()
             passmgr.add_password(*authinfo)
 
             handlers.extend((httpbasicauthhandler(passmgr),
                              httpdigestauthhandler(passmgr)))
 
         o = urllib2.build_opener(*handlers)
         o.addheaders = [('User-Agent', 'git/1.7.8.0')]  # fake some git
 
         q = {"service": 'git-upload-pack'}
         qs = '?%s' % urllib.urlencode(q)
         cu = "%s%s" % (test_uri, qs)
         req = urllib2.Request(cu, None, {})
 
         try:
             resp = o.open(req)
             return resp.code == 200
         except Exception, e:
             # means it cannot be cloned
             raise urllib2.URLError("[%s] %s" % (url, e))
 
     def _get_repo(self, create, src_url=None, update_after_clone=False,
                   bare=False):
         if create and os.path.exists(self.path):
             raise RepositoryError("Location already exist")
         if src_url and not create:
             raise RepositoryError("Create should be set to True if src_url is "
                                   "given (clone operation creates repository)")
         try:
             if create and src_url:
                 GitRepository._check_url(src_url)
                 self.clone(src_url, update_after_clone, bare)
                 return Repo(self.path)
             elif create:
                 os.mkdir(self.path)
                 if bare:
                     return Repo.init_bare(self.path)
                 else:
                     return Repo.init(self.path)
             else:
                 return Repo(self.path)
         except (NotGitRepository, OSError), err:
             raise RepositoryError(err)
 
     def _get_all_revisions(self):
         # we must check if this repo is not empty, since later command
         # fails if it is. And it's cheaper to ask than throw the subprocess
         # errors
         try:
             self._repo.head()
         except KeyError:
             return []
         cmd = 'rev-list --all --reverse --date-order'
         try:
             so, se = self.run_git_command(cmd)
         except RepositoryError:
             # Can be raised for empty repositories
             return []
         return so.splitlines()
 
     def _get_all_revisions2(self):
         #alternate implementation using dulwich
         includes = [x[1][0] for x in self._parsed_refs.iteritems()
                     if x[1][1] != 'T']
         return [c.commit.id for c in self._repo.get_walker(include=includes)]
 
     def _get_revision(self, revision):
         """
         For git backend we always return integer here. This way we ensure
         that changset's revision attribute would become integer.
         """
         pattern = re.compile(r'^[[0-9a-fA-F]{12}|[0-9a-fA-F]{40}]$')
         is_bstr = lambda o: isinstance(o, (str, unicode))
         is_null = lambda o: len(o) == revision.count('0')
 
         if len(self.revisions) == 0:
             raise EmptyRepositoryError("There are no changesets yet")
 
         if revision in (None, '', 'tip', 'HEAD', 'head', -1):
             revision = self.revisions[-1]
 
         if ((is_bstr(revision) and revision.isdigit() and len(revision) < 12)
             or isinstance(revision, int) or is_null(revision)):
             try:
                 revision = self.revisions[int(revision)]
             except:
                 raise ChangesetDoesNotExistError("Revision %r does not exist "
                     "for this repository %s" % (revision, self))
 
         elif is_bstr(revision):
             # get by branch/tag name
             _ref_revision = self._parsed_refs.get(revision)
             _tags_shas = self.tags.values()
             if _ref_revision:  # and _ref_revision[1] in ['H', 'RH', 'T']:
                 return _ref_revision[0]
 
             # maybe it's a tag ? we don't have them in self.revisions
             elif revision in _tags_shas:
                 return _tags_shas[_tags_shas.index(revision)]
 
             elif not pattern.match(revision) or revision not in self.revisions:
                 raise ChangesetDoesNotExistError("Revision %r does not exist "
                     "for this repository %s" % (revision, self))
 
         # Ensure we return full id
         if not pattern.match(str(revision)):
             raise ChangesetDoesNotExistError("Given revision %r not recognized"
                 % revision)
         return revision
 
     def _get_archives(self, archive_name='tip'):
 
         for i in [('zip', '.zip'), ('gz', '.tar.gz'), ('bz2', '.tar.bz2')]:
                 yield {"type": i[0], "extension": i[1], "node": archive_name}
 
     def _get_url(self, url):
         """
         Returns normalized url. If schema is not given, would fall to
         filesystem (``file:///``) schema.
         """
         url = str(url)
         if url != 'default' and not '://' in url:
             url = ':///'.join(('file', url))
         return url
 
     @LazyProperty
     def name(self):
         return os.path.basename(self.path)
 
     @LazyProperty
     def last_change(self):
         """
         Returns last change made on this repository as datetime object
         """
         return date_fromtimestamp(self._get_mtime(), makedate()[1])
 
     def _get_mtime(self):
         try:
             return time.mktime(self.get_changeset().date.timetuple())
         except RepositoryError:
             idx_loc = '' if self.bare else '.git'
             # fallback to filesystem
             in_path = os.path.join(self.path, idx_loc, "index")
             he_path = os.path.join(self.path, idx_loc, "HEAD")
             if os.path.exists(in_path):
                 return os.stat(in_path).st_mtime
             else:
                 return os.stat(he_path).st_mtime
 
     @LazyProperty
     def description(self):
         idx_loc = '' if self.bare else '.git'
         undefined_description = u'unknown'
         description_path = os.path.join(self.path, idx_loc, 'description')
         if os.path.isfile(description_path):
             return safe_unicode(open(description_path).read())
         else:
             return undefined_description
 
     @LazyProperty
     def contact(self):
         undefined_contact = u'Unknown'
         return undefined_contact
 
     @property
     def branches(self):
         if not self.revisions:
             return {}
         sortkey = lambda ctx: ctx[0]
         _branches = [(x[0], x[1][0])
                      for x in self._parsed_refs.iteritems() if x[1][1] == 'H']
         return OrderedDict(sorted(_branches, key=sortkey, reverse=False))
 
     @LazyProperty
     def tags(self):
         return self._get_tags()
 
     def _get_tags(self):
         if not self.revisions:
             return {}
 
         sortkey = lambda ctx: ctx[0]
         _tags = [(x[0], x[1][0])
                  for x in self._parsed_refs.iteritems() if x[1][1] == 'T']
         return OrderedDict(sorted(_tags, key=sortkey, reverse=True))
 
     def tag(self, name, user, revision=None, message=None, date=None,
             **kwargs):
         """
         Creates and returns a tag for the given ``revision``.
 
         :param name: name for new tag
         :param user: full username, i.e.: "Joe Doe <[e-mail]>"
         :param revision: changeset id for which new tag would be created
         :param message: message of the tag's commit
         :param date: date of tag's commit
 
         :raises TagAlreadyExistError: if tag with same name already exists
         """
         if name in self.tags:
             raise TagAlreadyExistError("Tag %s already exists" % name)
         changeset = self.get_changeset(revision)
         message = message or "Added tag %s for commit %s" % (name,
             changeset.raw_id)
         self._repo.refs["refs/tags/%s" % name] = changeset._commit.id
 
         self._parsed_refs = self._get_parsed_refs()
         self.tags = self._get_tags()
         return changeset
 
     def remove_tag(self, name, user, message=None, date=None):
         """
         Removes tag with the given ``name``.
 
         :param name: name of the tag to be removed
         :param user: full username, i.e.: "Joe Doe <[e-mail]>"
         :param message: message of the tag's removal commit
         :param date: date of tag's removal commit
 
         :raises TagDoesNotExistError: if tag with given name does not exists
         """
         if name not in self.tags:
             raise TagDoesNotExistError("Tag %s does not exist" % name)
         tagpath = posixpath.join(self._repo.refs.path, 'refs', 'tags', name)
         try:
             os.remove(tagpath)
             self._parsed_refs = self._get_parsed_refs()
             self.tags = self._get_tags()
         except OSError, e:
             raise RepositoryError(e.strerror)
 
     @LazyProperty
     def _parsed_refs(self):
         return self._get_parsed_refs()
 
     def _get_parsed_refs(self):
         refs = self._repo.get_refs()
         keys = [('refs/heads/', 'H'),
                 ('refs/remotes/origin/', 'RH'),
                 ('refs/tags/', 'T')]
         _refs = {}
         for ref, sha in refs.iteritems():
             for k, type_ in keys:
                 if ref.startswith(k):
                     _key = ref[len(k):]
                     if type_ == 'T':
                         obj = self._repo.get_object(sha)
                         if isinstance(obj, Tag):
                             sha = self._repo.get_object(sha).object[1]
                     _refs[_key] = [sha, type_]
                     break
         return _refs
 
     def _heads(self, reverse=False):
         refs = self._repo.get_refs()
         heads = {}
 
         for key, val in refs.items():
             for ref_key in ['refs/heads/', 'refs/remotes/origin/']:
                 if key.startswith(ref_key):
                     n = key[len(ref_key):]
                     if n not in ['HEAD']:
                         heads[n] = val
 
         return heads if reverse else dict((y, x) for x, y in heads.iteritems())
 
     def get_changeset(self, revision=None):
         """
         Returns ``GitChangeset`` object representing commit from git repository
         at the given revision or head (most recent commit) if None given.
         """
         if isinstance(revision, GitChangeset):
             return revision
         revision = self._get_revision(revision)
         changeset = GitChangeset(repository=self, revision=revision)
         return changeset
 
     def get_changesets(self, start=None, end=None, start_date=None,
            end_date=None, branch_name=None, reverse=False):
         """
         Returns iterator of ``GitChangeset`` objects from start to end (both
         are inclusive), in ascending date order (unless ``reverse`` is set).
 
         :param start: changeset ID, as str; first returned changeset
         :param end: changeset ID, as str; last returned changeset
         :param start_date: if specified, changesets with commit date less than
           ``start_date`` would be filtered out from returned set
         :param end_date: if specified, changesets with commit date greater than
           ``end_date`` would be filtered out from returned set
         :param branch_name: if specified, changesets not reachable from given
           branch would be filtered out from returned set
         :param reverse: if ``True``, returned generator would be reversed
           (meaning that returned changesets would have descending date order)
 
         :raise BranchDoesNotExistError: If given ``branch_name`` does not
             exist.
         :raise ChangesetDoesNotExistError: If changeset for given ``start`` or
           ``end`` could not be found.
 
         """
         if branch_name and branch_name not in self.branches:
             raise BranchDoesNotExistError("Branch '%s' not found" \
                                           % branch_name)
         # %H at format means (full) commit hash, initial hashes are retrieved
         # in ascending date order
         cmd_template = 'log --date-order --reverse --pretty=format:"%H"'
         cmd_params = {}
         if start_date:
             cmd_template += ' --since "$since"'
             cmd_params['since'] = start_date.strftime('%m/%d/%y %H:%M:%S')
         if end_date:
             cmd_template += ' --until "$until"'
             cmd_params['until'] = end_date.strftime('%m/%d/%y %H:%M:%S')
         if branch_name:
             cmd_template += ' $branch_name'
             cmd_params['branch_name'] = branch_name
         else:
             cmd_template += ' --all'
 
         cmd = Template(cmd_template).safe_substitute(**cmd_params)
         revs = self.run_git_command(cmd)[0].splitlines()
         start_pos = 0
         end_pos = len(revs)
         if start:
             _start = self._get_revision(start)
             try:
                 start_pos = revs.index(_start)
             except ValueError:
                 pass
 
         if end is not None:
             _end = self._get_revision(end)
             try:
                 end_pos = revs.index(_end)
             except ValueError:
                 pass
 
         if None not in [start, end] and start_pos > end_pos:
             raise RepositoryError('start cannot be after end')
 
         if end_pos is not None:
             end_pos += 1
 
         revs = revs[start_pos:end_pos]
         if reverse:
             revs = reversed(revs)
         for rev in revs:
             yield self.get_changeset(rev)
 
     def get_diff(self, rev1, rev2, path=None, ignore_whitespace=False,
                  context=3):
         """
         Returns (git like) *diff*, as plain text. Shows changes introduced by
         ``rev2`` since ``rev1``.
 
         :param rev1: Entry point from which diff is shown. Can be
           ``self.EMPTY_CHANGESET`` - in this case, patch showing all
           the changes since empty state of the repository until ``rev2``
         :param rev2: Until which revision changes should be shown.
         :param ignore_whitespace: If set to ``True``, would not show whitespace
           changes. Defaults to ``False``.
         :param context: How many lines before/after changed lines should be
           shown. Defaults to ``3``.
         """
         flags = ['-U%s' % context, '--full-index', '--binary', '-p', '-M', '--abbrev=40']
         if ignore_whitespace:
             flags.append('-w')
 
         if hasattr(rev1, 'raw_id'):
             rev1 = getattr(rev1, 'raw_id')
 
         if hasattr(rev2, 'raw_id'):
             rev2 = getattr(rev2, 'raw_id')
 
         if rev1 == self.EMPTY_CHANGESET:
             rev2 = self.get_changeset(rev2).raw_id
             cmd = ' '.join(['show'] + flags + [rev2])
         else:
             rev1 = self.get_changeset(rev1).raw_id
             rev2 = self.get_changeset(rev2).raw_id
             cmd = ' '.join(['diff'] + flags + [rev1, rev2])
 
         if path:
             cmd += ' -- "%s"' % path
 
         stdout, stderr = self.run_git_command(cmd)
         # If we used 'show' command, strip first few lines (until actual diff
         # starts)
         if rev1 == self.EMPTY_CHANGESET:
             lines = stdout.splitlines()
             x = 0
             for line in lines:
                 if line.startswith('diff'):
                     break
                 x += 1
             # Append new line just like 'diff' command do
             stdout = '\n'.join(lines[x:]) + '\n'
         return stdout
 
     @LazyProperty
     def in_memory_changeset(self):
         """
         Returns ``GitInMemoryChangeset`` object for this repository.
         """
         return GitInMemoryChangeset(self)
 
     def clone(self, url, update_after_clone=True, bare=False):
         """
         Tries to clone changes from external location.
 
         :param update_after_clone: If set to ``False``, git won't checkout
           working directory
         :param bare: If set to ``True``, repository would be cloned into
           *bare* git repository (no working directory at all).
         """
         url = self._get_url(url)
         cmd = ['clone']
         if bare:
             cmd.append('--bare')
         elif not update_after_clone:
             cmd.append('--no-checkout')
         cmd += ['--', '"%s"' % url, '"%s"' % self.path]
         cmd = ' '.join(cmd)
         # If error occurs run_git_command raises RepositoryError already
         self.run_git_command(cmd)
 
     def pull(self, url):
         """
         Tries to pull changes from external location.
         """
         url = self._get_url(url)
         cmd = ['pull']
         cmd.append("--ff-only")
         cmd.append(url)
         cmd = ' '.join(cmd)
         # If error occurs run_git_command raises RepositoryError already
         self.run_git_command(cmd)
 
     def fetch(self, url):
         """
         Tries to pull changes from external location.
         """
         url = self._get_url(url)
         so, se = self.run_git_command('ls-remote -h %s' % url)
         refs = []
         for line in (x for x in so.splitlines()):
             sha, ref = line.split('\t')
             refs.append(ref)
         refs = ' '.join(('+%s:%s' % (r, r) for r in refs))
         cmd = '''fetch %s -- %s''' % (url, refs)
         self.run_git_command(cmd)
 
     @LazyProperty
     def workdir(self):
         """
         Returns ``Workdir`` instance for this repository.
         """
         return GitWorkdir(self)
 
     def get_config_value(self, section, name, config_file=None):
         """
         Returns configuration value for a given [``section``] and ``name``.
 
         :param section: Section we want to retrieve value from
         :param name: Name of configuration we want to retrieve
         :param config_file: A path to file which should be used to retrieve
           configuration from (might also be a list of file paths)
         """
         if config_file is None:
             config_file = []
         elif isinstance(config_file, basestring):
             config_file = [config_file]
 
         def gen_configs():
             for path in config_file + self._config_files:
                 try:
                     yield ConfigFile.from_path(path)
                 except (IOError, OSError, ValueError):
                     continue
 
         for config in gen_configs():
             try:
                 return config.get(section, name)
             except KeyError:
                 continue
         return None
 
     def get_user_name(self, config_file=None):
         """
         Returns user's name from global configuration file.
 
         :param config_file: A path to file which should be used to retrieve
           configuration from (might also be a list of file paths)
         """
         return self.get_config_value('user', 'name', config_file)
 
     def get_user_email(self, config_file=None):
         """
         Returns user's email from global configuration file.
 
         :param config_file: A path to file which should be used to retrieve
           configuration from (might also be a list of file paths)
         """
         return self.get_config_value('user', 'email', config_file)