convert: use raw string for regular expressions

This avoids a SyntaxWarning on Python 3.8.

Differential Revision: https://phab.mercurial-scm.org/D5821

hgext/convert/monotone.py

@@ -1,373 +1,373 b''
 # monotone.py - monotone support for the convert extension
 #
 #  Copyright 2008, 2009 Mikkel Fahnoe Jorgensen <[e-mail]> and
 #  others
 #
 # This software may be used and distributed according to the terms of the
 # GNU General Public License version 2 or any later version.
 from __future__ import absolute_import
 
 import os
 import re
 
 from mercurial.i18n import _
 from mercurial import (
     error,
     pycompat,
 )
 from mercurial.utils import dateutil
 
 from . import common
 
 class monotone_source(common.converter_source, common.commandline):
     def __init__(self, ui, repotype, path=None, revs=None):
         common.converter_source.__init__(self, ui, repotype, path, revs)
         if revs and len(revs) > 1:
             raise error.Abort(_('monotone source does not support specifying '
                                'multiple revs'))
         common.commandline.__init__(self, ui, 'mtn')
 
         self.ui = ui
         self.path = path
         self.automatestdio = False
         self.revs = revs
 
         norepo = common.NoRepo(_("%s does not look like a monotone repository")
                              % path)
         if not os.path.exists(os.path.join(path, '_MTN')):
             # Could be a monotone repository (SQLite db file)
             try:
                 f = open(path, 'rb')
                 header = f.read(16)
                 f.close()
             except IOError:
                 header = ''
             if header != 'SQLite format 3\x00':
                 raise norepo
 
         # regular expressions for parsing monotone output
         space    = br'\s*'
         name     = br'\s+"((?:\\"|[^"])*)"\s*'
         value    = name
         revision = br'\s+\[(\w+)\]\s*'
         lines    = br'(?:.|\n)+'
 
         self.dir_re      = re.compile(space + "dir" + name)
         self.file_re     = re.compile(space + "file" + name +
                                       "content" + revision)
         self.add_file_re = re.compile(space + "add_file" + name +
                                       "content" + revision)
         self.patch_re    = re.compile(space + "patch" + name +
                                       "from" + revision + "to" + revision)
         self.rename_re   = re.compile(space + "rename" + name + "to" + name)
         self.delete_re   = re.compile(space + "delete" + name)
         self.tag_re      = re.compile(space + "tag" + name + "revision" +
                                       revision)
         self.cert_re     = re.compile(lines + space + "name" + name +
                                       "value" + value)
 
         attr = space + "file" + lines + space + "attr" + space
         self.attr_execute_re = re.compile(attr  + '"mtn:execute"' +
                                           space + '"true"')
 
         # cached data
         self.manifest_rev = None
         self.manifest = None
         self.files = None
         self.dirs  = None
 
         common.checktool('mtn', abort=False)
 
     def mtnrun(self, *args, **kwargs):
         if self.automatestdio:
             return self.mtnrunstdio(*args, **kwargs)
         else:
             return self.mtnrunsingle(*args, **kwargs)
 
     def mtnrunsingle(self, *args, **kwargs):
         kwargs[r'd'] = self.path
         return self.run0('automate', *args, **kwargs)
 
     def mtnrunstdio(self, *args, **kwargs):
         # Prepare the command in automate stdio format
         kwargs = pycompat.byteskwargs(kwargs)
         command = []
         for k, v in kwargs.iteritems():
             command.append("%d:%s" % (len(k), k))
             if v:
                 command.append("%d:%s" % (len(v), v))
         if command:
             command.insert(0, 'o')
             command.append('e')
 
         command.append('l')
         for arg in args:
             command.append("%d:%s" % (len(arg), arg))
         command.append('e')
         command = ''.join(command)
 
         self.ui.debug("mtn: sending '%s'\n" % command)
         self.mtnwritefp.write(command)
         self.mtnwritefp.flush()
 
         return self.mtnstdioreadcommandoutput(command)
 
     def mtnstdioreadpacket(self):
         read = None
         commandnbr = ''
         while read != ':':
             read = self.mtnreadfp.read(1)
             if not read:
                 raise error.Abort(_('bad mtn packet - no end of commandnbr'))
             commandnbr += read
         commandnbr = commandnbr[:-1]
 
         stream = self.mtnreadfp.read(1)
         if stream not in 'mewptl':
             raise error.Abort(_('bad mtn packet - bad stream type %s') % stream)
 
         read = self.mtnreadfp.read(1)
         if read != ':':
             raise error.Abort(_('bad mtn packet - no divider before size'))
 
         read = None
         lengthstr = ''
         while read != ':':
             read = self.mtnreadfp.read(1)
             if not read:
                 raise error.Abort(_('bad mtn packet - no end of packet size'))
             lengthstr += read
         try:
             length = pycompat.long(lengthstr[:-1])
         except TypeError:
             raise error.Abort(_('bad mtn packet - bad packet size %s')
                 % lengthstr)
 
         read = self.mtnreadfp.read(length)
         if len(read) != length:
             raise error.Abort(_("bad mtn packet - unable to read full packet "
                 "read %s of %s") % (len(read), length))
 
         return (commandnbr, stream, length, read)
 
     def mtnstdioreadcommandoutput(self, command):
         retval = []
         while True:
             commandnbr, stream, length, output = self.mtnstdioreadpacket()
             self.ui.debug('mtn: read packet %s:%s:%s\n' %
                 (commandnbr, stream, length))
 
             if stream == 'l':
                 # End of command
                 if output != '0':
                     raise error.Abort(_("mtn command '%s' returned %s") %
                         (command, output))
                 break
             elif stream in 'ew':
                 # Error, warning output
                 self.ui.warn(_('%s error:\n') % self.command)
                 self.ui.warn(output)
             elif stream == 'p':
                 # Progress messages
                 self.ui.debug('mtn: ' + output)
             elif stream == 'm':
                 # Main stream - command output
                 retval.append(output)
 
         return ''.join(retval)
 
     def mtnloadmanifest(self, rev):
         if self.manifest_rev == rev:
             return
         self.manifest = self.mtnrun("get_manifest_of", rev).split("\n\n")
         self.manifest_rev = rev
         self.files = {}
         self.dirs = {}
 
         for e in self.manifest:
             m = self.file_re.match(e)
             if m:
                 attr = ""
                 name = m.group(1)
                 node = m.group(2)
                 if self.attr_execute_re.match(e):
                     attr += "x"
                 self.files[name] = (node, attr)
             m = self.dir_re.match(e)
             if m:
                 self.dirs[m.group(1)] = True
 
     def mtnisfile(self, name, rev):
         # a non-file could be a directory or a deleted or renamed file
         self.mtnloadmanifest(rev)
         return name in self.files
 
     def mtnisdir(self, name, rev):
         self.mtnloadmanifest(rev)
         return name in self.dirs
 
     def mtngetcerts(self, rev):
         certs = {"author":"<missing>", "date":"<missing>",
             "changelog":"<missing>", "branch":"<missing>"}
         certlist = self.mtnrun("certs", rev)
         # mtn < 0.45:
         #   key "[e-mail]"
         # mtn >= 0.45:
         #   key [ff58a7ffb771907c4ff68995eada1c4da068d328]
-        certlist = re.split('\n\n      key ["\[]', certlist)
+        certlist = re.split(br'\n\n      key ["\[]', certlist)
         for e in certlist:
             m = self.cert_re.match(e)
             if m:
                 name, value = m.groups()
                 value = value.replace(r'\"', '"')
                 value = value.replace(r'\\', '\\')
                 certs[name] = value
         # Monotone may have subsecond dates: 2005-02-05T09:39:12.364306
         # and all times are stored in UTC
         certs["date"] = certs["date"].split('.')[0] + " UTC"
         return certs
 
     # implement the converter_source interface:
 
     def getheads(self):
         if not self.revs:
             return self.mtnrun("leaves").splitlines()
         else:
             return self.revs
 
     def getchanges(self, rev, full):
         if full:
             raise error.Abort(_("convert from monotone does not support "
                               "--full"))
         revision = self.mtnrun("get_revision", rev).split("\n\n")
         files = {}
         ignoremove = {}
         renameddirs = []
         copies = {}
         for e in revision:
             m = self.add_file_re.match(e)
             if m:
                 files[m.group(1)] = rev
                 ignoremove[m.group(1)] = rev
             m = self.patch_re.match(e)
             if m:
                 files[m.group(1)] = rev
             # Delete/rename is handled later when the convert engine
             # discovers an IOError exception from getfile,
             # but only if we add the "from" file to the list of changes.
             m = self.delete_re.match(e)
             if m:
                 files[m.group(1)] = rev
             m = self.rename_re.match(e)
             if m:
                 toname = m.group(2)
                 fromname = m.group(1)
                 if self.mtnisfile(toname, rev):
                     ignoremove[toname] = 1
                     copies[toname] = fromname
                     files[toname] = rev
                     files[fromname] = rev
                 elif self.mtnisdir(toname, rev):
                     renameddirs.append((fromname, toname))
 
         # Directory renames can be handled only once we have recorded
         # all new files
         for fromdir, todir in renameddirs:
             renamed = {}
             for tofile in self.files:
                 if tofile in ignoremove:
                     continue
                 if tofile.startswith(todir + '/'):
                     renamed[tofile] = fromdir + tofile[len(todir):]
                     # Avoid chained moves like:
                     # d1(/a) => d3/d1(/a)
                     # d2 => d3
                     ignoremove[tofile] = 1
             for tofile, fromfile in renamed.items():
                 self.ui.debug (_("copying file in renamed directory "
                                  "from '%s' to '%s'")
                                % (fromfile, tofile), '\n')
                 files[tofile] = rev
                 copies[tofile] = fromfile
             for fromfile in renamed.values():
                 files[fromfile] = rev
 
         return (files.items(), copies, set())
 
     def getfile(self, name, rev):
         if not self.mtnisfile(name, rev):
             return None, None
         try:
             data = self.mtnrun("get_file_of", name, r=rev)
         except Exception:
             return None, None
         self.mtnloadmanifest(rev)
         node, attr = self.files.get(name, (None, ""))
         return data, attr
 
     def getcommit(self, rev):
         extra = {}
         certs = self.mtngetcerts(rev)
         if certs.get('suspend') == certs["branch"]:
             extra['close'] = 1
         dateformat = "%Y-%m-%dT%H:%M:%S"
         return common.commit(
             author=certs["author"],
             date=dateutil.datestr(dateutil.strdate(certs["date"], dateformat)),
             desc=certs["changelog"],
             rev=rev,
             parents=self.mtnrun("parents", rev).splitlines(),
             branch=certs["branch"],
             extra=extra)
 
     def gettags(self):
         tags = {}
         for e in self.mtnrun("tags").split("\n\n"):
             m = self.tag_re.match(e)
             if m:
                 tags[m.group(1)] = m.group(2)
         return tags
 
     def getchangedfiles(self, rev, i):
         # This function is only needed to support --filemap
         # ... and we don't support that
         raise NotImplementedError
 
     def before(self):
         # Check if we have a new enough version to use automate stdio
         try:
             versionstr = self.mtnrunsingle("interface_version")
             version = float(versionstr)
         except Exception:
             raise error.Abort(_("unable to determine mtn automate interface "
                 "version"))
 
         if version >= 12.0:
             self.automatestdio = True
             self.ui.debug("mtn automate version %f - using automate stdio\n" %
                 version)
 
             # launch the long-running automate stdio process
             self.mtnwritefp, self.mtnreadfp = self._run2('automate', 'stdio',
                 '-d', self.path)
             # read the headers
             read = self.mtnreadfp.readline()
             if read != 'format-version: 2\n':
                 raise error.Abort(_('mtn automate stdio header unexpected: %s')
                     % read)
             while read != '\n':
                 read = self.mtnreadfp.readline()
                 if not read:
                     raise error.Abort(_("failed to reach end of mtn automate "
                         "stdio headers"))
         else:
             self.ui.debug("mtn automate version %s - not using automate stdio "
                 "(automate >= 12.0 - mtn >= 0.46 is needed)\n" % version)
 
     def after(self):
         if self.automatestdio:
             self.mtnwritefp.close()
             self.mtnwritefp = None
             self.mtnreadfp.close()
             self.mtnreadfp = None
 

hgext/convert/p4.py

@@ -1,378 +1,378 b''
 # Perforce source for convert extension.
 #
 # Copyright 2009, Frank Kingswood <[e-mail]>
 #
 # This software may be used and distributed according to the terms of the
 # GNU General Public License version 2 or any later version.
 from __future__ import absolute_import
 
 import marshal
 import re
 
 from mercurial.i18n import _
 from mercurial import (
     error,
     util,
 )
 from mercurial.utils import (
     dateutil,
     procutil,
     stringutil,
 )
 
 from . import common
 
 def loaditer(f):
     "Yield the dictionary objects generated by p4"
     try:
         while True:
             d = marshal.load(f)
             if not d:
                 break
             yield d
     except EOFError:
         pass
 
 def decodefilename(filename):
     """Perforce escapes special characters @, #, *, or %
     with %40, %23, %2A, or %25 respectively
 
     >>> decodefilename(b'portable-net45%252Bnetcore45%252Bwp8%252BMonoAndroid')
     'portable-net45%2Bnetcore45%2Bwp8%2BMonoAndroid'
     >>> decodefilename(b'//Depot/Directory/%2525/%2523/%23%40.%2A')
     '//Depot/Directory/%25/%23/#@.*'
     """
     replacements = [('%2A', '*'), ('%23', '#'), ('%40', '@'), ('%25', '%')]
     for k, v in replacements:
         filename = filename.replace(k, v)
     return filename
 
 class p4_source(common.converter_source):
     def __init__(self, ui, repotype, path, revs=None):
         # avoid import cycle
         from . import convcmd
 
         super(p4_source, self).__init__(ui, repotype, path, revs=revs)
 
         if "/" in path and not path.startswith('//'):
             raise common.NoRepo(_('%s does not look like a P4 repository') %
                                 path)
 
         common.checktool('p4', abort=False)
 
         self.revmap = {}
         self.encoding = self.ui.config('convert', 'p4.encoding',
                                        convcmd.orig_encoding)
         self.re_type = re.compile(
-            "([a-z]+)?(text|binary|symlink|apple|resource|unicode|utf\d+)"
-            "(\+\w+)?$")
+            br"([a-z]+)?(text|binary|symlink|apple|resource|unicode|utf\d+)"
+            br"(\+\w+)?$")
         self.re_keywords = re.compile(
-            r"\$(Id|Header|Date|DateTime|Change|File|Revision|Author)"
-            r":[^$\n]*\$")
-        self.re_keywords_old = re.compile("\$(Id|Header):[^$\n]*\$")
+            br"\$(Id|Header|Date|DateTime|Change|File|Revision|Author)"
+            br":[^$\n]*\$")
+        self.re_keywords_old = re.compile(br"\$(Id|Header):[^$\n]*\$")
 
         if revs and len(revs) > 1:
             raise error.Abort(_("p4 source does not support specifying "
                                "multiple revisions"))
 
     def setrevmap(self, revmap):
         """Sets the parsed revmap dictionary.
 
         Revmap stores mappings from a source revision to a target revision.
         It is set in convertcmd.convert and provided by the user as a file
         on the commandline.
 
         Revisions in the map are considered beeing present in the
         repository and ignored during _parse(). This allows for incremental
         imports if a revmap is provided.
         """
         self.revmap = revmap
 
     def _parse_view(self, path):
         "Read changes affecting the path"
         cmd = 'p4 -G changes -s submitted %s' % procutil.shellquote(path)
         stdout = procutil.popen(cmd, mode='rb')
         p4changes = {}
         for d in loaditer(stdout):
             c = d.get("change", None)
             if c:
                 p4changes[c] = True
         return p4changes
 
     def _parse(self, ui, path):
         "Prepare list of P4 filenames and revisions to import"
         p4changes = {}
         changeset = {}
         files_map = {}
         copies_map = {}
         localname = {}
         depotname = {}
         heads = []
 
         ui.status(_('reading p4 views\n'))
 
         # read client spec or view
         if "/" in path:
             p4changes.update(self._parse_view(path))
             if path.startswith("//") and path.endswith("/..."):
                 views = {path[:-3]:""}
             else:
                 views = {"//": ""}
         else:
             cmd = 'p4 -G client -o %s' % procutil.shellquote(path)
             clientspec = marshal.load(procutil.popen(cmd, mode='rb'))
 
             views = {}
             for client in clientspec:
                 if client.startswith("View"):
                     sview, cview = clientspec[client].split()
                     p4changes.update(self._parse_view(sview))
                     if sview.endswith("...") and cview.endswith("..."):
                         sview = sview[:-3]
                         cview = cview[:-3]
                     cview = cview[2:]
                     cview = cview[cview.find("/") + 1:]
                     views[sview] = cview
 
         # list of changes that affect our source files
         p4changes = p4changes.keys()
         p4changes.sort(key=int)
 
         # list with depot pathnames, longest first
         vieworder = views.keys()
         vieworder.sort(key=len, reverse=True)
 
         # handle revision limiting
         startrev = self.ui.config('convert', 'p4.startrev')
 
         # now read the full changelists to get the list of file revisions
         ui.status(_('collecting p4 changelists\n'))
         lastid = None
         for change in p4changes:
             if startrev and int(change) < int(startrev):
                 continue
             if self.revs and int(change) > int(self.revs[0]):
                 continue
             if change in self.revmap:
                 # Ignore already present revisions, but set the parent pointer.
                 lastid = change
                 continue
 
             if lastid:
                 parents = [lastid]
             else:
                 parents = []
 
             d = self._fetch_revision(change)
             c = self._construct_commit(d, parents)
 
             descarr = c.desc.splitlines(True)
             if len(descarr) > 0:
                 shortdesc = descarr[0].rstrip('\r\n')
             else:
                 shortdesc = '**empty changelist description**'
 
             t = '%s %s' % (c.rev, repr(shortdesc)[1:-1])
             ui.status(stringutil.ellipsis(t, 80) + '\n')
 
             files = []
             copies = {}
             copiedfiles = []
             i = 0
             while ("depotFile%d" % i) in d and ("rev%d" % i) in d:
                 oldname = d["depotFile%d" % i]
                 filename = None
                 for v in vieworder:
                     if oldname.lower().startswith(v.lower()):
                         filename = decodefilename(views[v] + oldname[len(v):])
                         break
                 if filename:
                     files.append((filename, d["rev%d" % i]))
                     depotname[filename] = oldname
                     if (d.get("action%d" % i) == "move/add"):
                         copiedfiles.append(filename)
                     localname[oldname] = filename
                 i += 1
 
             # Collect information about copied files
             for filename in copiedfiles:
                 oldname = depotname[filename]
 
                 flcmd = 'p4 -G filelog %s' \
                       % procutil.shellquote(oldname)
                 flstdout = procutil.popen(flcmd, mode='rb')
 
                 copiedfilename = None
                 for d in loaditer(flstdout):
                     copiedoldname = None
 
                     i = 0
                     while ("change%d" % i) in d:
                         if (d["change%d" % i] == change and
                             d["action%d" % i] == "move/add"):
                             j = 0
                             while ("file%d,%d" % (i, j)) in d:
                                 if d["how%d,%d" % (i, j)] == "moved from":
                                     copiedoldname = d["file%d,%d" % (i, j)]
                                     break
                                 j += 1
                         i += 1
 
                     if copiedoldname and copiedoldname in localname:
                         copiedfilename = localname[copiedoldname]
                         break
 
                 if copiedfilename:
                     copies[filename] = copiedfilename
                 else:
                     ui.warn(_("cannot find source for copied file: %s@%s\n")
                             % (filename, change))
 
             changeset[change] = c
             files_map[change] = files
             copies_map[change] = copies
             lastid = change
 
         if lastid and len(changeset) > 0:
             heads = [lastid]
 
         return {
             'changeset': changeset,
             'files': files_map,
             'copies': copies_map,
             'heads': heads,
             'depotname': depotname,
         }
 
     @util.propertycache
     def _parse_once(self):
         return self._parse(self.ui, self.path)
 
     @util.propertycache
     def copies(self):
         return self._parse_once['copies']
 
     @util.propertycache
     def files(self):
         return self._parse_once['files']
 
     @util.propertycache
     def changeset(self):
         return self._parse_once['changeset']
 
     @util.propertycache
     def heads(self):
         return self._parse_once['heads']
 
     @util.propertycache
     def depotname(self):
         return self._parse_once['depotname']
 
     def getheads(self):
         return self.heads
 
     def getfile(self, name, rev):
         cmd = 'p4 -G print %s' \
             % procutil.shellquote("%s#%s" % (self.depotname[name], rev))
 
         lasterror = None
         while True:
             stdout = procutil.popen(cmd, mode='rb')
 
             mode = None
             contents = []
             keywords = None
 
             for d in loaditer(stdout):
                 code = d["code"]
                 data = d.get("data")
 
                 if code == "error":
                     # if this is the first time error happened
                     # re-attempt getting the file
                     if not lasterror:
                         lasterror = IOError(d["generic"], data)
                         # this will exit inner-most for-loop
                         break
                     else:
                         raise lasterror
 
                 elif code == "stat":
                     action = d.get("action")
                     if action in ["purge", "delete", "move/delete"]:
                         return None, None
                     p4type = self.re_type.match(d["type"])
                     if p4type:
                         mode = ""
                         flags = ((p4type.group(1) or "")
                                + (p4type.group(3) or ""))
                         if "x" in flags:
                             mode = "x"
                         if p4type.group(2) == "symlink":
                             mode = "l"
                         if "ko" in flags:
                             keywords = self.re_keywords_old
                         elif "k" in flags:
                             keywords = self.re_keywords
 
                 elif code == "text" or code == "binary":
                     contents.append(data)
 
                 lasterror = None
 
             if not lasterror:
                 break
 
         if mode is None:
             return None, None
 
         contents = ''.join(contents)
 
         if keywords:
             contents = keywords.sub("$\\1$", contents)
         if mode == "l" and contents.endswith("\n"):
             contents = contents[:-1]
 
         return contents, mode
 
     def getchanges(self, rev, full):
         if full:
             raise error.Abort(_("convert from p4 does not support --full"))
         return self.files[rev], self.copies[rev], set()
 
     def _construct_commit(self, obj, parents=None):
         """
         Constructs a common.commit object from an unmarshalled
         `p4 describe` output
         """
         desc = self.recode(obj.get("desc", ""))
         date = (int(obj["time"]), 0)     # timezone not set
         if parents is None:
             parents = []
 
         return common.commit(author=self.recode(obj["user"]),
             date=dateutil.datestr(date, '%Y-%m-%d %H:%M:%S %1%2'),
             parents=parents, desc=desc, branch=None, rev=obj['change'],
             extra={"p4": obj['change'], "convert_revision": obj['change']})
 
     def _fetch_revision(self, rev):
         """Return an output of `p4 describe` including author, commit date as
         a dictionary."""
         cmd = "p4 -G describe -s %s" % rev
         stdout = procutil.popen(cmd, mode='rb')
         return marshal.load(stdout)
 
     def getcommit(self, rev):
         if rev in self.changeset:
             return self.changeset[rev]
         elif rev in self.revmap:
             d = self._fetch_revision(rev)
             return self._construct_commit(d, parents=None)
         raise error.Abort(
             _("cannot find %s in the revmap or parsed changesets") % rev)
 
     def gettags(self):
         return {}
 
     def getchangedfiles(self, rev, i):
         return sorted([x[0] for x in self.files[rev]])