##// END OF EJS Templates
Make patch.diff filelog cache LRU of 20 files. Fixes issue1738....
Make patch.diff filelog cache LRU of 20 files. Fixes issue1738. 20 files is as fast as 200 for hg diff -r 28015:30103 of mozilla-central. Ideally we'd use util.lrucachefunc, but the interface doesn't quite work.

File last commit:

r7190:aecea693 default
r9123:360f61c2 default
Show More
parsers.c
435 lines | 10.1 KiB | text/x-c | CLexer
Bryan O'Sullivan
manifest: improve parsing performance by 8x via a new C extension
r6389 /*
parsers.c - efficient content parsing
Copyright 2008 Matt Mackall <mpm@selenic.com> and others
This software may be used and distributed according to the terms of
the GNU General Public License, incorporated herein by reference.
*/
#include <Python.h>
#include <ctype.h>
#include <string.h>
static int hexdigit(char c)
{
if (c >= '0' && c <= '9')
return c - '0';
Matt Mackall
parsers: speed up hex decoding for manifests
r7092 if (c >= 'a' && c <= 'f')
return c - 'a' + 10;
Bryan O'Sullivan
manifest: improve parsing performance by 8x via a new C extension
r6389 if (c >= 'A' && c <= 'F')
return c - 'A' + 10;
Matt Mackall
parsers: speed up hex decoding for manifests
r7092 PyErr_SetString(PyExc_ValueError, "input contains non-hex character");
return 0;
Bryan O'Sullivan
manifest: improve parsing performance by 8x via a new C extension
r6389 }
/*
* Turn a hex-encoded string into binary.
*/
static PyObject *unhexlify(const char *str, int len)
{
Matt Mackall
parsers: speed up hex decoding for manifests
r7092 PyObject *ret;
Benoit Boissinot
fix const annotation warning
r6395 const char *c;
char *d;
Bryan O'Sullivan
manifest: improve parsing performance by 8x via a new C extension
r6389
ret = PyString_FromStringAndSize(NULL, len / 2);
if (!ret)
Matt Mackall
parsers: speed up hex decoding for manifests
r7092 return NULL;
Bryan O'Sullivan
manifest: improve parsing performance by 8x via a new C extension
r6389
Matt Mackall
parsers: speed up hex decoding for manifests
r7092 d = PyString_AS_STRING(ret);
Bryan O'Sullivan
manifest: improve parsing performance by 8x via a new C extension
r6389 for (c = str; c < str + len;) {
int hi = hexdigit(*c++);
int lo = hexdigit(*c++);
*d++ = (hi << 4) | lo;
}
Matt Mackall
parsers: clean up whitespace
r7091
Bryan O'Sullivan
manifest: improve parsing performance by 8x via a new C extension
r6389 return ret;
}
/*
* This code assumes that a manifest is stitched together with newline
* ('\n') characters.
*/
static PyObject *parse_manifest(PyObject *self, PyObject *args)
{
PyObject *mfdict, *fdict;
char *str, *cur, *start, *zero;
int len;
if (!PyArg_ParseTuple(args, "O!O!s#:parse_manifest",
&PyDict_Type, &mfdict,
&PyDict_Type, &fdict,
&str, &len))
goto quit;
for (start = cur = str, zero = NULL; cur < str + len; cur++) {
PyObject *file = NULL, *node = NULL;
PyObject *flags = NULL;
int nlen;
if (!*cur) {
zero = cur;
continue;
}
else if (*cur != '\n')
continue;
if (!zero) {
PyErr_SetString(PyExc_ValueError,
"manifest entry has no separator");
goto quit;
}
file = PyString_FromStringAndSize(start, zero - start);
if (!file)
goto bail;
nlen = cur - zero - 1;
node = unhexlify(zero + 1, nlen > 40 ? 40 : nlen);
if (!node)
goto bail;
if (nlen > 40) {
PyObject *flags;
flags = PyString_FromStringAndSize(zero + 41,
nlen - 40);
if (!flags)
goto bail;
if (PyDict_SetItem(fdict, file, flags) == -1)
goto bail;
}
if (PyDict_SetItem(mfdict, file, node) == -1)
goto bail;
start = cur + 1;
zero = NULL;
Py_XDECREF(flags);
Py_XDECREF(node);
Py_XDECREF(file);
continue;
bail:
Py_XDECREF(flags);
Py_XDECREF(node);
Py_XDECREF(file);
goto quit;
}
if (len > 0 && *(cur - 1) != '\n') {
PyErr_SetString(PyExc_ValueError,
"manifest contains trailing garbage");
goto quit;
}
Py_INCREF(Py_None);
return Py_None;
quit:
return NULL;
}
Matt Mackall
dirstate: C parsing extension
r7093 #ifdef _WIN32
# ifdef _MSC_VER
/* msvc 6.0 has problems */
# define inline __inline
typedef unsigned long uint32_t;
Dhruva Krishnamurthy
Fix missing uint64_t definition in parsers.c under Windows
r7122 typedef unsigned __int64 uint64_t;
Matt Mackall
dirstate: C parsing extension
r7093 # else
# include <stdint.h>
# endif
static uint32_t ntohl(uint32_t x)
{
return ((x & 0x000000ffUL) << 24) |
Thomas Arendsen Hein
Some additional space/tab cleanups
r7190 ((x & 0x0000ff00UL) << 8) |
((x & 0x00ff0000UL) >> 8) |
((x & 0xff000000UL) >> 24);
Matt Mackall
dirstate: C parsing extension
r7093 }
#else
/* not windows */
# include <sys/types.h>
# if defined __BEOS__ && !defined __HAIKU__
# include <ByteOrder.h>
# else
# include <arpa/inet.h>
# endif
# include <inttypes.h>
#endif
static PyObject *parse_dirstate(PyObject *self, PyObject *args)
{
PyObject *dmap, *cmap, *parents = NULL, *ret = NULL;
PyObject *fname = NULL, *cname = NULL, *entry = NULL;
char *str, *cur, *end, *cpos;
Benoit Boissinot
parsers.c: fix integer overflows...
r7174 int state, mode, size, mtime;
unsigned int flen;
Matt Mackall
dirstate: C parsing extension
r7093 int len;
char decode[16]; /* for alignment */
if (!PyArg_ParseTuple(args, "O!O!s#:parse_dirstate",
&PyDict_Type, &dmap,
&PyDict_Type, &cmap,
&str, &len))
goto quit;
/* read parents */
if (len < 40)
goto quit;
parents = Py_BuildValue("s#s#", str, 20, str + 20, 20);
if (!parents)
goto quit;
/* read filenames */
cur = str + 40;
end = str + len;
while (cur < end - 17) {
/* unpack header */
state = *cur;
memcpy(decode, cur + 1, 16);
mode = ntohl(*(uint32_t *)(decode));
size = ntohl(*(uint32_t *)(decode + 4));
mtime = ntohl(*(uint32_t *)(decode + 8));
flen = ntohl(*(uint32_t *)(decode + 12));
cur += 17;
Benoit Boissinot
parsers.c: fix integer overflows...
r7174 if (flen > end - cur) {
PyErr_SetString(PyExc_ValueError, "overflow in dirstate");
Matt Mackall
dirstate: C parsing extension
r7093 goto quit;
Benoit Boissinot
parsers.c: fix integer overflows...
r7174 }
Matt Mackall
dirstate: C parsing extension
r7093
entry = Py_BuildValue("ciii", state, mode, size, mtime);
if (!entry)
goto quit;
Benoit Boissinot
parsers.c: do not try to untrack after a failure
r7175 PyObject_GC_UnTrack(entry); /* don't waste time with this */
Matt Mackall
dirstate: C parsing extension
r7093
cpos = memchr(cur, 0, flen);
if (cpos) {
fname = PyString_FromStringAndSize(cur, cpos - cur);
cname = PyString_FromStringAndSize(cpos + 1,
flen - (cpos - cur) - 1);
if (!fname || !cname ||
PyDict_SetItem(cmap, fname, cname) == -1 ||
PyDict_SetItem(dmap, fname, entry) == -1)
goto quit;
Py_DECREF(cname);
} else {
fname = PyString_FromStringAndSize(cur, flen);
if (!fname ||
PyDict_SetItem(dmap, fname, entry) == -1)
goto quit;
}
cur += flen;
Py_DECREF(fname);
Py_DECREF(entry);
fname = cname = entry = NULL;
}
ret = parents;
Py_INCREF(ret);
quit:
Py_XDECREF(fname);
Py_XDECREF(cname);
Py_XDECREF(entry);
Py_XDECREF(parents);
return ret;
}
Bernhard Leiner
C implementation of revlog index parsing
r7108 const char nullid[20];
const int nullrev = -1;
Benoit Boissinot
index parser: fix refcounting in case of errors, refactor...
r7154 /* create an index tuple, insert into the nodemap */
static PyObject * _build_idx_entry(PyObject *nodemap, int n, uint64_t offset_flags,
int comp_len, int uncomp_len, int base_rev,
int link_rev, int parent_1, int parent_2,
const char *c_node_id)
{
int err;
PyObject *entry, *node_id, *n_obj;
node_id = PyString_FromStringAndSize(c_node_id, 20);
n_obj = PyInt_FromLong(n);
if (!node_id || !n_obj)
err = -1;
else
err = PyDict_SetItem(nodemap, node_id, n_obj);
Py_XDECREF(n_obj);
if (err)
goto error_dealloc;
entry = Py_BuildValue("LiiiiiiN", offset_flags, comp_len,
uncomp_len, base_rev, link_rev,
parent_1, parent_2, node_id);
if (!entry)
goto error_dealloc;
PyObject_GC_UnTrack(entry); /* don't waste time with this */
return entry;
error_dealloc:
Py_XDECREF(node_id);
return NULL;
}
Bernhard Leiner
C implementation of revlog index parsing
r7108 /* RevlogNG format (all in big endian, data may be inlined):
* 6 bytes: offset
* 2 bytes: flags
* 4 bytes: compressed length
* 4 bytes: uncompressed length
* 4 bytes: base revision
* 4 bytes: link revision
* 4 bytes: parent 1 revision
* 4 bytes: parent 2 revision
* 32 bytes: nodeid (only 20 bytes used)
*/
static int _parse_index_ng (const char *data, int size, int inlined,
PyObject *index, PyObject *nodemap)
{
Benoit Boissinot
index parser: fix refcounting in case of errors, refactor...
r7154 PyObject *entry;
int n = 0, err;
uint64_t offset_flags;
Bernhard Leiner
C implementation of revlog index parsing
r7108 int comp_len, uncomp_len, base_rev, link_rev, parent_1, parent_2;
Benoit Boissinot
index parser: fix refcounting in case of errors, refactor...
r7154 const char *c_node_id;
Bernhard Leiner
C implementation of revlog index parsing
r7108 const char *end = data + size;
Thomas Arendsen Hein
Copy index before parsing to enforce alignment with inline data present....
r7169 char decode[64]; /* to enforce alignment with inline data */
Bernhard Leiner
C implementation of revlog index parsing
r7108
while (data < end) {
Benoit Boissinot
parsers.c: fix integer overflows...
r7174 unsigned int step;
Thomas Arendsen Hein
Some additional space/tab cleanups
r7190
Thomas Arendsen Hein
Copy index before parsing to enforce alignment with inline data present....
r7169 memcpy(decode, data, 64);
Dirkjan Ochtman
clean up trailing spaces, leading spaces in C
r7186 offset_flags = ntohl(*((uint32_t *) (decode + 4)));
if (n == 0) /* mask out version number for the first entry */
offset_flags &= 0xFFFF;
else {
Thomas Arendsen Hein
Copy index before parsing to enforce alignment with inline data present....
r7169 uint32_t offset_high = ntohl(*((uint32_t *) decode));
Dirkjan Ochtman
clean up trailing spaces, leading spaces in C
r7186 offset_flags |= ((uint64_t) offset_high) << 32;
Benoit Boissinot
revlog parser: use ntohl() instead of ntohll() (fix endianness issues)
r7133 }
Thomas Arendsen Hein
Copy index before parsing to enforce alignment with inline data present....
r7169 comp_len = ntohl(*((uint32_t *) (decode + 8)));
uncomp_len = ntohl(*((uint32_t *) (decode + 12)));
base_rev = ntohl(*((uint32_t *) (decode + 16)));
link_rev = ntohl(*((uint32_t *) (decode + 20)));
parent_1 = ntohl(*((uint32_t *) (decode + 24)));
parent_2 = ntohl(*((uint32_t *) (decode + 28)));
c_node_id = decode + 32;
Benoit Boissinot
index parser: fix refcounting in case of errors, refactor...
r7154
entry = _build_idx_entry(nodemap, n, offset_flags,
comp_len, uncomp_len, base_rev,
link_rev, parent_1, parent_2,
c_node_id);
Bernhard Leiner
C implementation of revlog index parsing
r7108 if (!entry)
Benoit Boissinot
index parser: fix refcounting in case of errors, refactor...
r7154 return 0;
Bernhard Leiner
C implementation of revlog index parsing
r7108
if (inlined) {
Benoit Boissinot
index parser: fix refcounting in case of errors, refactor...
r7154 err = PyList_Append(index, entry);
Bernhard Leiner
C implementation of revlog index parsing
r7108 Py_DECREF(entry);
Benoit Boissinot
index parser: fix refcounting in case of errors, refactor...
r7154 if (err)
return 0;
} else
Bernhard Leiner
C implementation of revlog index parsing
r7108 PyList_SET_ITEM(index, n, entry); /* steals reference */
n++;
Benoit Boissinot
parsers.c: fix integer overflows...
r7174 step = 64 + (inlined ? comp_len : 0);
if (end - data < step)
break;
data += step;
Bernhard Leiner
C implementation of revlog index parsing
r7108 }
Benoit Boissinot
parsers.c: fix integer overflows...
r7174 if (data != end) {
Bernhard Leiner
C implementation of revlog index parsing
r7108 if (!PyErr_Occurred())
PyErr_SetString(PyExc_ValueError, "corrupt index file");
Benoit Boissinot
index parser: fix refcounting in case of errors, refactor...
r7154 return 0;
Bernhard Leiner
C implementation of revlog index parsing
r7108 }
Thomas Arendsen Hein
Some additional space/tab cleanups
r7190 /* create the nullid/nullrev entry in the nodemap and the
Bernhard Leiner
C implementation of revlog index parsing
r7108 * magic nullid entry in the index at [-1] */
Thomas Arendsen Hein
Some additional space/tab cleanups
r7190 entry = _build_idx_entry(nodemap,
Benoit Boissinot
index parser: fix refcounting in case of errors, refactor...
r7154 nullrev, 0, 0, 0, -1, -1, -1, -1, nullid);
Bernhard Leiner
C implementation of revlog index parsing
r7108 if (!entry)
Benoit Boissinot
index parser: fix refcounting in case of errors, refactor...
r7154 return 0;
Bernhard Leiner
C implementation of revlog index parsing
r7108 if (inlined) {
Benoit Boissinot
index parser: fix refcounting in case of errors, refactor...
r7154 err = PyList_Append(index, entry);
Bernhard Leiner
C implementation of revlog index parsing
r7108 Py_DECREF(entry);
Benoit Boissinot
index parser: fix refcounting in case of errors, refactor...
r7154 if (err)
return 0;
} else
Bernhard Leiner
C implementation of revlog index parsing
r7108 PyList_SET_ITEM(index, n, entry); /* steals reference */
return 1;
}
/* This function parses a index file and returns a Python tuple of the
* following format: (index, nodemap, cache)
*
* index: a list of tuples containing the RevlogNG records
* nodemap: a dict mapping node ids to indices in the index list
* cache: if data is inlined, a tuple (index_file_content, 0) else None
*/
static PyObject *parse_index(PyObject *self, PyObject *args)
{
const char *data;
int size, inlined;
Thomas Arendsen Hein
Some additional space/tab cleanups
r7190 PyObject *rval = NULL, *index = NULL, *nodemap = NULL, *cache = NULL;
Bernhard Leiner
C implementation of revlog index parsing
r7108 PyObject *data_obj = NULL, *inlined_obj;
if (!PyArg_ParseTuple(args, "s#O", &data, &size, &inlined_obj))
return NULL;
inlined = inlined_obj && PyObject_IsTrue(inlined_obj);
Thomas Arendsen Hein
Some additional space/tab cleanups
r7190 /* If no data is inlined, we know the size of the index list in
* advance: size divided by size of one one revlog record (64 bytes)
* plus one for the nullid */
Bernhard Leiner
C implementation of revlog index parsing
r7108 index = inlined ? PyList_New(0) : PyList_New(size / 64 + 1);
if (!index)
goto quit;
nodemap = PyDict_New();
Benoit Boissinot
index parser: fix refcounting in case of errors, refactor...
r7154 if (!nodemap)
goto quit;
Bernhard Leiner
C implementation of revlog index parsing
r7108
/* set up the cache return value */
if (inlined) {
/* Note that the reference to data_obj is only borrowed */
data_obj = PyTuple_GET_ITEM(args, 0);
cache = Py_BuildValue("iO", 0, data_obj);
if (!cache)
goto quit;
} else {
cache = Py_None;
Py_INCREF(Py_None);
}
/* actually populate the index and the nodemap with data */
if (!_parse_index_ng (data, size, inlined, index, nodemap))
goto quit;
rval = Py_BuildValue("NNN", index, nodemap, cache);
if (!rval)
goto quit;
return rval;
quit:
Py_XDECREF(index);
Py_XDECREF(nodemap);
Py_XDECREF(cache);
Py_XDECREF(rval);
return NULL;
}
Bryan O'Sullivan
manifest: improve parsing performance by 8x via a new C extension
r6389 static char parsers_doc[] = "Efficient content parsing.";
static PyMethodDef methods[] = {
{"parse_manifest", parse_manifest, METH_VARARGS, "parse a manifest\n"},
Matt Mackall
dirstate: C parsing extension
r7093 {"parse_dirstate", parse_dirstate, METH_VARARGS, "parse a dirstate\n"},
Bernhard Leiner
C implementation of revlog index parsing
r7108 {"parse_index", parse_index, METH_VARARGS, "parse a revlog index\n"},
Bryan O'Sullivan
manifest: improve parsing performance by 8x via a new C extension
r6389 {NULL, NULL}
};
PyMODINIT_FUNC initparsers(void)
{
Py_InitModule3("parsers", methods, parsers_doc);
}