##// END OF EJS Templates
wireprotov2: add phases to "changesetdata" command...
wireprotov2: add phases to "changesetdata" command This commit teaches the "changesetdata" wire protocol command to emit the phase state for each changeset. This is a different approach from existing phase transfer in a few ways. Previously, if there are no new revisions (or we're not using bundle2), we perform a "listkeys" request to retrieve phase heads. And when revision data is being transferred with bundle2, phases data is encoded in a standalone bundle2 part. In both cases, phases data is logically decoupled from the changeset data and is encountered/applied after changeset revision data is received. The new wire protocol purposefully tries to more tightly associate changeset metadata (phases, bookmarks, obsolescence markers, etc) with the changeset revision and index data itself, rather than have it live as a separate entity that must be fetched and processed separately. I reckon that one reason we didn't do this before was it was difficult to add new data types/fields without breaking existing consumers. By using CBOR maps to transfer changeset data and putting clients in control of what fields are requested / present in those maps, we can easily add additional changeset data while maintaining backwards compatibility. I believe this to be a superior approach to the problem. That being said, for performance reasons, we may need to resort to alternative mechanisms for transferring data like phases. But for now, I think giving the wire protocol the ability to transfer changeset metadata next to the changeset itself is a powerful feature because it is a raw, changeset-centric data API. And if you build simple APIs for accessing the fundamental units of repository data, you enable client-side experimentation (partial clone, etc). If it turns out that we need specialized APIs or mechanisms for transferring data like phases, we can build in those APIs later. For now, I'd like to see how far we can get on simple APIs. It's worth noting that when phase data is being requested, the server will also emit changeset records for nodes in the bases specified by the "noderange" argument. This is to ensure that phase-only updates for nodes the client has are available to the client, even if no new changesets will be transferred. Differential Revision: https://phab.mercurial-scm.org/D4483

File last commit:

r39484:ad76032d stable
r39668:c1aacb0d default
Show More
bdiff.c
339 lines | 7.4 KiB | text/x-c | CLexer
Yuya Nishihara
bdiff: switch to policy importer...
r32369 /*
bdiff.c - efficient binary diff extension for Mercurial
Copyright 2005, 2006 Matt Mackall <mpm@selenic.com>
This software may be used and distributed according to the terms of
the GNU General Public License, incorporated herein by reference.
Based roughly on Python difflib
*/
#define PY_SSIZE_T_CLEAN
#include <Python.h>
Gregory Szorc
cext: reorder #include...
r34439 #include <limits.h>
Yuya Nishihara
bdiff: switch to policy importer...
r32369 #include <stdlib.h>
#include <string.h>
#include "bdiff.h"
#include "bitmanipulation.h"
Jun Wu
bdiff: add a xdiffblocks method...
r36693 #include "thirdparty/xdiff/xdiff.h"
Yuya Nishihara
bdiff: switch to policy importer...
r32369 #include "util.h"
static PyObject *blocks(PyObject *self, PyObject *args)
{
PyObject *sa, *sb, *rl = NULL, *m;
struct bdiff_line *a, *b;
struct bdiff_hunk l, *h;
int an, bn, count, pos = 0;
l.next = NULL;
if (!PyArg_ParseTuple(args, "SS:bdiff", &sa, &sb))
return NULL;
an = bdiff_splitlines(PyBytes_AsString(sa), PyBytes_Size(sa), &a);
bn = bdiff_splitlines(PyBytes_AsString(sb), PyBytes_Size(sb), &b);
if (!a || !b)
goto nomem;
count = bdiff_diff(a, an, b, bn, &l);
if (count < 0)
goto nomem;
rl = PyList_New(count);
if (!rl)
goto nomem;
for (h = l.next; h; h = h->next) {
m = Py_BuildValue("iiii", h->a1, h->a2, h->b1, h->b2);
PyList_SetItem(rl, pos, m);
pos++;
}
nomem:
free(a);
free(b);
bdiff_freehunks(l.next);
return rl ? rl : PyErr_NoMemory();
}
static PyObject *bdiff(PyObject *self, PyObject *args)
{
Gregory Szorc
cext: accept arguments as Py_buffer...
r36673 Py_buffer ba, bb;
char *rb, *ia, *ib;
Yuya Nishihara
bdiff: switch to policy importer...
r32369 PyObject *result = NULL;
Gregory Szorc
cext: refactor cleanup code in bdiff()...
r36672 struct bdiff_line *al = NULL, *bl = NULL;
Yuya Nishihara
bdiff: switch to policy importer...
r32369 struct bdiff_hunk l, *h;
int an, bn, count;
Py_ssize_t len = 0, la, lb, li = 0, lcommon = 0, lmax;
Gregory Szorc
cext: refactor cleanup code in bdiff()...
r36672 PyThreadState *_save = NULL;
Yuya Nishihara
bdiff: switch to policy importer...
r32369
l.next = NULL;
Gregory Szorc
cext: accept arguments as Py_buffer...
r36673 if (!PyArg_ParseTuple(args, PY23("s*s*:bdiff", "y*y*:bdiff"), &ba, &bb))
Yuya Nishihara
bdiff: switch to policy importer...
r32369 return NULL;
Gregory Szorc
cext: accept arguments as Py_buffer...
r36673 if (!PyBuffer_IsContiguous(&ba, 'C') || ba.ndim > 1) {
PyErr_SetString(PyExc_ValueError, "bdiff input not contiguous");
goto cleanup;
}
if (!PyBuffer_IsContiguous(&bb, 'C') || bb.ndim > 1) {
PyErr_SetString(PyExc_ValueError, "bdiff input not contiguous");
goto cleanup;
}
la = ba.len;
lb = bb.len;
Yuya Nishihara
bdiff: switch to policy importer...
r32369 if (la > UINT_MAX || lb > UINT_MAX) {
PyErr_SetString(PyExc_ValueError, "bdiff inputs too large");
Gregory Szorc
cext: accept arguments as Py_buffer...
r36673 goto cleanup;
Yuya Nishihara
bdiff: switch to policy importer...
r32369 }
_save = PyEval_SaveThread();
lmax = la > lb ? lb : la;
Gregory Szorc
cext: accept arguments as Py_buffer...
r36673 for (ia = ba.buf, ib = bb.buf; li < lmax && *ia == *ib;
++li, ++ia, ++ib) {
Yuya Nishihara
bdiff: switch to policy importer...
r32369 if (*ia == '\n')
lcommon = li + 1;
Gregory Szorc
cext: accept arguments as Py_buffer...
r36673 }
Yuya Nishihara
bdiff: switch to policy importer...
r32369 /* we can almost add: if (li == lmax) lcommon = li; */
Matt Harbison
bdiff: avoid pointer arithmetic on void*...
r36699 an = bdiff_splitlines((char *)ba.buf + lcommon, la - lcommon, &al);
bn = bdiff_splitlines((char *)bb.buf + lcommon, lb - lcommon, &bl);
Gregory Szorc
cext: refactor cleanup code in bdiff()...
r36672 if (!al || !bl) {
PyErr_NoMemory();
goto cleanup;
}
Yuya Nishihara
bdiff: switch to policy importer...
r32369
count = bdiff_diff(al, an, bl, bn, &l);
Gregory Szorc
cext: refactor cleanup code in bdiff()...
r36672 if (count < 0) {
PyErr_NoMemory();
goto cleanup;
}
Yuya Nishihara
bdiff: switch to policy importer...
r32369
/* calculate length of output */
la = lb = 0;
for (h = l.next; h; h = h->next) {
if (h->a1 != la || h->b1 != lb)
len += 12 + bl[h->b1].l - bl[lb].l;
la = h->a2;
lb = h->b2;
}
PyEval_RestoreThread(_save);
_save = NULL;
result = PyBytes_FromStringAndSize(NULL, len);
if (!result)
Gregory Szorc
cext: refactor cleanup code in bdiff()...
r36672 goto cleanup;
Yuya Nishihara
bdiff: switch to policy importer...
r32369
/* build binary patch */
rb = PyBytes_AsString(result);
la = lb = 0;
for (h = l.next; h; h = h->next) {
if (h->a1 != la || h->b1 != lb) {
len = bl[h->b1].l - bl[lb].l;
putbe32((uint32_t)(al[la].l + lcommon - al->l), rb);
Augie Fackler
bdiff: add to clang-format oversight...
r36072 putbe32((uint32_t)(al[h->a1].l + lcommon - al->l),
rb + 4);
Yuya Nishihara
bdiff: switch to policy importer...
r32369 putbe32((uint32_t)len, rb + 8);
memcpy(rb + 12, bl[lb].l, len);
rb += 12 + len;
}
la = h->a2;
lb = h->b2;
}
Gregory Szorc
cext: refactor cleanup code in bdiff()...
r36672 cleanup:
Yuya Nishihara
bdiff: switch to policy importer...
r32369 if (_save)
PyEval_RestoreThread(_save);
Gregory Szorc
cext: accept arguments as Py_buffer...
r36673 PyBuffer_Release(&ba);
PyBuffer_Release(&bb);
Josef 'Jeff' Sipek
cext: stop worrying and love the free(NULL)...
r38320 free(al);
free(bl);
Yuya Nishihara
bdiff: one more safe call of bdiff_freehunks(NULL)
r38328 bdiff_freehunks(l.next);
Gregory Szorc
cext: refactor cleanup code in bdiff()...
r36672 return result;
Yuya Nishihara
bdiff: switch to policy importer...
r32369 }
/*
* If allws != 0, remove all whitespace (' ', \t and \r). Otherwise,
* reduce whitespace sequences to a single space and trim remaining whitespace
* from end of lines.
*/
static PyObject *fixws(PyObject *self, PyObject *args)
{
PyObject *s, *result = NULL;
char allws, c;
const char *r;
Py_ssize_t i, rlen, wlen = 0;
char *w;
if (!PyArg_ParseTuple(args, "Sb:fixws", &s, &allws))
return NULL;
r = PyBytes_AsString(s);
rlen = PyBytes_Size(s);
w = (char *)PyMem_Malloc(rlen ? rlen : 1);
if (!w)
goto nomem;
for (i = 0; i != rlen; i++) {
c = r[i];
if (c == ' ' || c == '\t' || c == '\r') {
if (!allws && (wlen == 0 || w[wlen - 1] != ' '))
w[wlen++] = ' ';
Augie Fackler
bdiff: add to clang-format oversight...
r36072 } else if (c == '\n' && !allws && wlen > 0 &&
w[wlen - 1] == ' ') {
Yuya Nishihara
bdiff: switch to policy importer...
r32369 w[wlen - 1] = '\n';
} else {
w[wlen++] = c;
}
}
result = PyBytes_FromStringAndSize(w, wlen);
nomem:
PyMem_Free(w);
return result ? result : PyErr_NoMemory();
}
Augie Fackler
bdiff: write a native version of splitnewlines...
r36163 static bool sliceintolist(PyObject *list, Py_ssize_t destidx,
const char *source, Py_ssize_t len)
{
PyObject *sliced = PyBytes_FromStringAndSize(source, len);
if (sliced == NULL)
return false;
PyList_SET_ITEM(list, destidx, sliced);
return true;
}
static PyObject *splitnewlines(PyObject *self, PyObject *args)
{
const char *text;
Py_ssize_t nelts = 0, size, i, start = 0;
PyObject *result = NULL;
Yuya Nishihara
py3: bulk-replace 'const char*' format specifier passed to PyArg_ParseTuple*()...
r36638 if (!PyArg_ParseTuple(args, PY23("s#", "y#"), &text, &size)) {
Augie Fackler
bdiff: write a native version of splitnewlines...
r36163 goto abort;
}
if (!size) {
return PyList_New(0);
}
/* This loops to size-1 because if the last byte is a newline,
* we don't want to perform a split there. */
for (i = 0; i < size - 1; ++i) {
if (text[i] == '\n') {
++nelts;
}
}
if ((result = PyList_New(nelts + 1)) == NULL)
goto abort;
nelts = 0;
for (i = 0; i < size - 1; ++i) {
if (text[i] == '\n') {
if (!sliceintolist(result, nelts++, text + start,
i - start + 1))
goto abort;
start = i + 1;
}
}
if (!sliceintolist(result, nelts++, text + start, size - start))
goto abort;
return result;
abort:
Py_XDECREF(result);
return NULL;
}
Matt Harbison
bdiff: convert more longs to int64_t...
r36934 static int hunk_consumer(int64_t a1, int64_t a2, int64_t b1, int64_t b2,
void *priv)
Jun Wu
bdiff: add a xdiffblocks method...
r36693 {
PyObject *rl = (PyObject *)priv;
Julien Cristau
bdiff: fix yet more fallout from xdiff long/int64 conversion (issue5885)...
r38016 PyObject *m = Py_BuildValue("LLLL", a1, a2, b1, b2);
Yuya Nishihara
xdiff: fix leak in hunk_consumer()...
r39484 int r;
Jun Wu
bdiff: add a xdiffblocks method...
r36693 if (!m)
return -1;
Yuya Nishihara
xdiff: fix leak in hunk_consumer()...
r39484 r = PyList_Append(rl, m);
Py_DECREF(m);
return r;
Jun Wu
bdiff: add a xdiffblocks method...
r36693 }
static PyObject *xdiffblocks(PyObject *self, PyObject *args)
{
Py_ssize_t la, lb;
mmfile_t a, b;
PyObject *rl;
xpparam_t xpp = {
XDF_INDENT_HEURISTIC, /* flags */
};
xdemitconf_t xecfg = {
XDL_EMIT_BDIFFHUNK, /* flags */
hunk_consumer, /* hunk_consume_func */
};
xdemitcb_t ecb = {
NULL, /* priv */
};
if (!PyArg_ParseTuple(args, PY23("s#s#", "y#y#"), &a.ptr, &la, &b.ptr,
&lb))
return NULL;
a.size = la;
b.size = lb;
rl = PyList_New(0);
if (!rl)
return PyErr_NoMemory();
ecb.priv = rl;
if (xdl_diff(&a, &b, &xpp, &xecfg, &ecb) != 0) {
Py_DECREF(rl);
return PyErr_NoMemory();
}
return rl;
}
Yuya Nishihara
bdiff: switch to policy importer...
r32369 static char mdiff_doc[] = "Efficient binary diff.";
static PyMethodDef methods[] = {
Augie Fackler
bdiff: add to clang-format oversight...
r36072 {"bdiff", bdiff, METH_VARARGS, "calculate a binary diff\n"},
{"blocks", blocks, METH_VARARGS, "find a list of matching lines\n"},
{"fixws", fixws, METH_VARARGS, "normalize diff whitespaces\n"},
Augie Fackler
bdiff: write a native version of splitnewlines...
r36163 {"splitnewlines", splitnewlines, METH_VARARGS,
"like str.splitlines, but only split on newlines\n"},
Jun Wu
bdiff: add a xdiffblocks method...
r36693 {"xdiffblocks", xdiffblocks, METH_VARARGS,
"find a list of matching lines using xdiff algorithm\n"},
Augie Fackler
bdiff: add to clang-format oversight...
r36072 {NULL, NULL},
Yuya Nishihara
bdiff: switch to policy importer...
r32369 };
Jun Wu
bdiff: add a xdiffblocks method...
r36693 static const int version = 3;
Yuya Nishihara
bdiff: switch to policy importer...
r32369
#ifdef IS_PY3K
static struct PyModuleDef bdiff_module = {
Augie Fackler
bdiff: add to clang-format oversight...
r36072 PyModuleDef_HEAD_INIT, "bdiff", mdiff_doc, -1, methods,
Yuya Nishihara
bdiff: switch to policy importer...
r32369 };
PyMODINIT_FUNC PyInit_bdiff(void)
{
PyObject *m;
m = PyModule_Create(&bdiff_module);
PyModule_AddIntConstant(m, "version", version);
return m;
}
#else
PyMODINIT_FUNC initbdiff(void)
{
PyObject *m;
m = Py_InitModule3("bdiff", methods, mdiff_doc);
PyModule_AddIntConstant(m, "version", version);
}
#endif