diff options
Diffstat (limited to 'builtin-apply.c')
| -rw-r--r-- | builtin-apply.c | 2303 |
1 files changed, 2303 insertions, 0 deletions
diff --git a/builtin-apply.c b/builtin-apply.c new file mode 100644 index 0000000000..f16c753bee --- /dev/null +++ b/builtin-apply.c @@ -0,0 +1,2303 @@ +/* + * apply.c + * + * Copyright (C) Linus Torvalds, 2005 + * + * This applies patches on top of some (arbitrary) version of the SCM. + * + */ +#include <fnmatch.h> +#include "cache.h" +#include "cache-tree.h" +#include "quote.h" +#include "blob.h" +#include "delta.h" +#include "builtin.h" + +// --check turns on checking that the working tree matches the +// files that are being modified, but doesn't apply the patch +// --stat does just a diffstat, and doesn't actually apply +// --numstat does numeric diffstat, and doesn't actually apply +// --index-info shows the old and new index info for paths if available. +// --index updates the cache as well. +// --cached updates only the cache without ever touching the working tree. +// +static const char *prefix; +static int prefix_length = -1; +static int newfd = -1; + +static int p_value = 1; +static int allow_binary_replacement = 0; +static int check_index = 0; +static int write_index = 0; +static int cached = 0; +static int diffstat = 0; +static int numstat = 0; +static int summary = 0; +static int check = 0; +static int apply = 1; +static int no_add = 0; +static int show_index_info = 0; +static int line_termination = '\n'; +static unsigned long p_context = -1; +static const char apply_usage[] = +"git-apply [--stat] [--numstat] [--summary] [--check] [--index] [--cached] [--apply] [--no-add] [--index-info] [--allow-binary-replacement] [-z] [-pNUM] [-CNUM] [--whitespace=<nowarn|warn|error|error-all|strip>] <patch>..."; + +static enum whitespace_eol { + nowarn_whitespace, + warn_on_whitespace, + error_on_whitespace, + strip_whitespace, +} new_whitespace = warn_on_whitespace; +static int whitespace_error = 0; +static int squelch_whitespace_errors = 5; +static int applied_after_stripping = 0; +static const char *patch_input_file = NULL; + +static void parse_whitespace_option(const char *option) +{ + if (!option) { + new_whitespace = warn_on_whitespace; + return; + } + if (!strcmp(option, "warn")) { + new_whitespace = warn_on_whitespace; + return; + } + if (!strcmp(option, "nowarn")) { + new_whitespace = nowarn_whitespace; + return; + } + if (!strcmp(option, "error")) { + new_whitespace = error_on_whitespace; + return; + } + if (!strcmp(option, "error-all")) { + new_whitespace = error_on_whitespace; + squelch_whitespace_errors = 0; + return; + } + if (!strcmp(option, "strip")) { + new_whitespace = strip_whitespace; + return; + } + die("unrecognized whitespace option '%s'", option); +} + +static void set_default_whitespace_mode(const char *whitespace_option) +{ + if (!whitespace_option && !apply_default_whitespace) { + new_whitespace = (apply + ? warn_on_whitespace + : nowarn_whitespace); + } +} + +/* + * For "diff-stat" like behaviour, we keep track of the biggest change + * we've seen, and the longest filename. That allows us to do simple + * scaling. + */ +static int max_change, max_len; + +/* + * Various "current state", notably line numbers and what + * file (and how) we're patching right now.. The "is_xxxx" + * things are flags, where -1 means "don't know yet". + */ +static int linenr = 1; + +struct fragment { + unsigned long leading, trailing; + unsigned long oldpos, oldlines; + unsigned long newpos, newlines; + const char *patch; + int size; + struct fragment *next; +}; + +struct patch { + char *new_name, *old_name, *def_name; + unsigned int old_mode, new_mode; + int is_rename, is_copy, is_new, is_delete, is_binary; +#define BINARY_DELTA_DEFLATED 1 +#define BINARY_LITERAL_DEFLATED 2 + unsigned long deflate_origlen; + int lines_added, lines_deleted; + int score; + struct fragment *fragments; + char *result; + unsigned long resultsize; + char old_sha1_prefix[41]; + char new_sha1_prefix[41]; + struct patch *next; +}; + +#define CHUNKSIZE (8192) +#define SLOP (16) + +static void *read_patch_file(int fd, unsigned long *sizep) +{ + unsigned long size = 0, alloc = CHUNKSIZE; + void *buffer = xmalloc(alloc); + + for (;;) { + int nr = alloc - size; + if (nr < 1024) { + alloc += CHUNKSIZE; + buffer = xrealloc(buffer, alloc); + nr = alloc - size; + } + nr = xread(fd, buffer + size, nr); + if (!nr) + break; + if (nr < 0) + die("git-apply: read returned %s", strerror(errno)); + size += nr; + } + *sizep = size; + + /* + * Make sure that we have some slop in the buffer + * so that we can do speculative "memcmp" etc, and + * see to it that it is NUL-filled. + */ + if (alloc < size + SLOP) + buffer = xrealloc(buffer, size + SLOP); + memset(buffer + size, 0, SLOP); + return buffer; +} + +static unsigned long linelen(const char *buffer, unsigned long size) +{ + unsigned long len = 0; + while (size--) { + len++; + if (*buffer++ == '\n') + break; + } + return len; +} + +static int is_dev_null(const char *str) +{ + return !memcmp("/dev/null", str, 9) && isspace(str[9]); +} + +#define TERM_SPACE 1 +#define TERM_TAB 2 + +static int name_terminate(const char *name, int namelen, int c, int terminate) +{ + if (c == ' ' && !(terminate & TERM_SPACE)) + return 0; + if (c == '\t' && !(terminate & TERM_TAB)) + return 0; + + return 1; +} + +static char * find_name(const char *line, char *def, int p_value, int terminate) +{ + int len; + const char *start = line; + char *name; + + if (*line == '"') { + /* Proposed "new-style" GNU patch/diff format; see + * http://marc.theaimsgroup.com/?l=git&m=112927316408690&w=2 + */ + name = unquote_c_style(line, NULL); + if (name) { + char *cp = name; + while (p_value) { + cp = strchr(name, '/'); + if (!cp) + break; + cp++; + p_value--; + } + if (cp) { + /* name can later be freed, so we need + * to memmove, not just return cp + */ + memmove(name, cp, strlen(cp) + 1); + free(def); + return name; + } + else { + free(name); + name = NULL; + } + } + } + + for (;;) { + char c = *line; + + if (isspace(c)) { + if (c == '\n') + break; + if (name_terminate(start, line-start, c, terminate)) + break; + } + line++; + if (c == '/' && !--p_value) + start = line; + } + if (!start) + return def; + len = line - start; + if (!len) + return def; + + /* + * Generally we prefer the shorter name, especially + * if the other one is just a variation of that with + * something else tacked on to the end (ie "file.orig" + * or "file~"). + */ + if (def) { + int deflen = strlen(def); + if (deflen < len && !strncmp(start, def, deflen)) + return def; + } + + name = xmalloc(len + 1); + memcpy(name, start, len); + name[len] = 0; + free(def); + return name; +} + +/* + * Get the name etc info from the --/+++ lines of a traditional patch header + * + * NOTE! This hardcodes "-p1" behaviour in filename detection. + * + * FIXME! The end-of-filename heuristics are kind of screwy. For existing + * files, we can happily check the index for a match, but for creating a + * new file we should try to match whatever "patch" does. I have no idea. + */ +static void parse_traditional_patch(const char *first, const char *second, struct patch *patch) +{ + char *name; + + first += 4; // skip "--- " + second += 4; // skip "+++ " + if (is_dev_null(first)) { + patch->is_new = 1; + patch->is_delete = 0; + name = find_name(second, NULL, p_value, TERM_SPACE | TERM_TAB); + patch->new_name = name; + } else if (is_dev_null(second)) { + patch->is_new = 0; + patch->is_delete = 1; + name = find_name(first, NULL, p_value, TERM_SPACE | TERM_TAB); + patch->old_name = name; + } else { + name = find_name(first, NULL, p_value, TERM_SPACE | TERM_TAB); + name = find_name(second, name, p_value, TERM_SPACE | TERM_TAB); + patch->old_name = patch->new_name = name; + } + if (!name) + die("unable to find filename in patch at line %d", linenr); +} + +static int gitdiff_hdrend(const char *line, struct patch *patch) +{ + return -1; +} + +/* + * We're anal about diff header consistency, to make + * sure that we don't end up having strange ambiguous + * patches floating around. + * + * As a result, gitdiff_{old|new}name() will check + * their names against any previous information, just + * to make sure.. + */ +static char *gitdiff_verify_name(const char *line, int isnull, char *orig_name, const char *oldnew) +{ + if (!orig_name && !isnull) + return find_name(line, NULL, 1, 0); + + if (orig_name) { + int len; + const char *name; + char *another; + name = orig_name; + len = strlen(name); + if (isnull) + die("git-apply: bad git-diff - expected /dev/null, got %s on line %d", name, linenr); + another = find_name(line, NULL, 1, 0); + if (!another || memcmp(another, name, len)) + die("git-apply: bad git-diff - inconsistent %s filename on line %d", oldnew, linenr); + free(another); + return orig_name; + } + else { + /* expect "/dev/null" */ + if (memcmp("/dev/null", line, 9) || line[9] != '\n') + die("git-apply: bad git-diff - expected /dev/null on line %d", linenr); + return NULL; + } +} + +static int gitdiff_oldname(const char *line, struct patch *patch) +{ + patch->old_name = gitdiff_verify_name(line, patch->is_new, patch->old_name, "old"); + return 0; +} + +static int gitdiff_newname(const char *line, struct patch *patch) +{ + patch->new_name = gitdiff_verify_name(line, patch->is_delete, patch->new_name, "new"); + return 0; +} + +static int gitdiff_oldmode(const char *line, struct patch *patch) +{ + patch->old_mode = strtoul(line, NULL, 8); + return 0; +} + +static int gitdiff_newmode(const char *line, struct patch *patch) +{ + patch->new_mode = strtoul(line, NULL, 8); + return 0; +} + +static int gitdiff_delete(const char *line, struct patch *patch) +{ + patch->is_delete = 1; + patch->old_name = patch->def_name; + return gitdiff_oldmode(line, patch); +} + +static int gitdiff_newfile(const char *line, struct patch *patch) +{ + patch->is_new = 1; + patch->new_name = patch->def_name; + return gitdiff_newmode(line, patch); +} + +static int gitdiff_copysrc(const char *line, struct patch *patch) +{ + patch->is_copy = 1; + patch->old_name = find_name(line, NULL, 0, 0); + return 0; +} + +static int gitdiff_copydst(const char *line, struct patch *patch) +{ + patch->is_copy = 1; + patch->new_name = find_name(line, NULL, 0, 0); + return 0; +} + +static int gitdiff_renamesrc(const char *line, struct patch *patch) +{ + patch->is_rename = 1; + patch->old_name = find_name(line, NULL, 0, 0); + return 0; +} + +static int gitdiff_renamedst(const char *line, struct patch *patch) +{ + patch->is_rename = 1; + patch->new_name = find_name(line, NULL, 0, 0); + return 0; +} + +static int gitdiff_similarity(const char *line, struct patch *patch) +{ + if ((patch->score = strtoul(line, NULL, 10)) == ULONG_MAX) + patch->score = 0; + return 0; +} + +static int gitdiff_dissimilarity(const char *line, struct patch *patch) +{ + if ((patch->score = strtoul(line, NULL, 10)) == ULONG_MAX) + patch->score = 0; + return 0; +} + +static int gitdiff_index(const char *line, struct patch *patch) +{ + /* index line is N hexadecimal, "..", N hexadecimal, + * and optional space with octal mode. + */ + const char *ptr, *eol; + int len; + + ptr = strchr(line, '.'); + if (!ptr || ptr[1] != '.' || 40 < ptr - line) + return 0; + len = ptr - line; + memcpy(patch->old_sha1_prefix, line, len); + patch->old_sha1_prefix[len] = 0; + + line = ptr + 2; + ptr = strchr(line, ' '); + eol = strchr(line, '\n'); + + if (!ptr || eol < ptr) + ptr = eol; + len = ptr - line; + + if (40 < len) + return 0; + memcpy(patch->new_sha1_prefix, line, len); + patch->new_sha1_prefix[len] = 0; + if (*ptr == ' ') + patch->new_mode = patch->old_mode = strtoul(ptr+1, NULL, 8); + return 0; +} + +/* + * This is normal for a diff that doesn't change anything: we'll fall through + * into the next diff. Tell the parser to break out. + */ +static int gitdiff_unrecognized(const char *line, struct patch *patch) +{ + return -1; +} + +static const char *stop_at_slash(const char *line, int llen) +{ + int i; + + for (i = 0; i < llen; i++) { + int ch = line[i]; + if (ch == '/') + return line + i; + } + return NULL; +} + +/* This is to extract the same name that appears on "diff --git" + * line. We do not find and return anything if it is a rename + * patch, and it is OK because we will find the name elsewhere. + * We need to reliably find name only when it is mode-change only, + * creation or deletion of an empty file. In any of these cases, + * both sides are the same name under a/ and b/ respectively. + */ +static char *git_header_name(char *line, int llen) +{ + int len; + const char *name; + const char *second = NULL; + + line += strlen("diff --git "); + llen -= strlen("diff --git "); + + if (*line == '"') { + const char *cp; + char *first = unquote_c_style(line, &second); + if (!first) + return NULL; + + /* advance to the first slash */ + cp = stop_at_slash(first, strlen(first)); + if (!cp || cp == first) { + /* we do not accept absolute paths */ + free_first_and_fail: + free(first); + return NULL; + } + len = strlen(cp+1); + memmove(first, cp+1, len+1); /* including NUL */ + + /* second points at one past closing dq of name. + * find the second name. + */ + while ((second < line + llen) && isspace(*second)) + second++; + + if (line + llen <= second) + goto free_first_and_fail; + if (*second == '"') { + char *sp = unquote_c_style(second, NULL); + if (!sp) + goto free_first_and_fail; + cp = stop_at_slash(sp, strlen(sp)); + if (!cp || cp == sp) { + free_both_and_fail: + free(sp); + goto free_first_and_fail; + } + /* They must match, otherwise ignore */ + if (strcmp(cp+1, first)) + goto free_both_and_fail; + free(sp); + return first; + } + + /* unquoted second */ + cp = stop_at_slash(second, line + llen - second); + if (!cp || cp == second) + goto free_first_and_fail; + cp++; + if (line + llen - cp != len + 1 || + memcmp(first, cp, len)) + goto free_first_and_fail; + return first; + } + + /* unquoted first name */ + name = stop_at_slash(line, llen); + if (!name || name == line) + return NULL; + + name++; + + /* since the first name is unquoted, a dq if exists must be + * the beginning of the second name. + */ + for (second = name; second < line + llen; second++) { + if (*second == '"') { + const char *cp = second; + const char *np; + char *sp = unquote_c_style(second, NULL); + + if (!sp) + return NULL; + np = stop_at_slash(sp, strlen(sp)); + if (!np || np == sp) { + free_second_and_fail: + free(sp); + return NULL; + } + np++; + len = strlen(np); + if (len < cp - name && + !strncmp(np, name, len) && + isspace(name[len])) { + /* Good */ + memmove(sp, np, len + 1); + return sp; + } + goto free_second_and_fail; + } + } + + /* + * Accept a name only if it shows up twice, exactly the same + * form. + */ + for (len = 0 ; ; len++) { + char c = name[len]; + + switch (c) { + default: + continue; + case '\n': + return NULL; + case '\t': case ' ': + second = name+len; + for (;;) { + char c = *second++; + if (c == '\n') + return NULL; + if (c == '/') + break; + } + if (second[len] == '\n' && !memcmp(name, second, len)) { + char *ret = xmalloc(len + 1); + memcpy(ret, name, len); + ret[len] = 0; + return ret; + } + } + } + return NULL; +} + +/* Verify that we recognize the lines following a git header */ +static int parse_git_header(char *line, int len, unsigned int size, struct patch *patch) +{ + unsigned long offset; + + /* A git diff has explicit new/delete information, so we don't guess */ + patch->is_new = 0; + patch->is_delete = 0; + + /* + * Some things may not have the old name in the + * rest of the headers anywhere (pure mode changes, + * or removing or adding empty files), so we get + * the default name from the header. + */ + patch->def_name = git_header_name(line, len); + + line += len; + size -= len; + linenr++; + for (offset = len ; size > 0 ; offset += len, size -= len, line += len, linenr++) { + static const struct opentry { + const char *str; + int (*fn)(const char *, struct patch *); + } optable[] = { + { "@@ -", gitdiff_hdrend }, + { "--- ", gitdiff_oldname }, + { "+++ ", gitdiff_newname }, + { "old mode ", gitdiff_oldmode }, + { "new mode ", gitdiff_newmode }, + { "deleted file mode ", gitdiff_delete }, + { "new file mode ", gitdiff_newfile }, + { "copy from ", gitdiff_copysrc }, + { "copy to ", gitdiff_copydst }, + { "rename old ", gitdiff_renamesrc }, + { "rename new ", gitdiff_renamedst }, + { "rename from ", gitdiff_renamesrc }, + { "rename to ", gitdiff_renamedst }, + { "similarity index ", gitdiff_similarity }, + { "dissimilarity index ", gitdiff_dissimilarity }, + { "index ", gitdiff_index }, + { "", gitdiff_unrecognized }, + }; + int i; + + len = linelen(line, size); + if (!len || line[len-1] != '\n') + break; + for (i = 0; i < ARRAY_SIZE(optable); i++) { + const struct opentry *p = optable + i; + int oplen = strlen(p->str); + if (len < oplen || memcmp(p->str, line, oplen)) + continue; + if (p->fn(line + oplen, patch) < 0) + return offset; + break; + } + } + + return offset; +} + +static int parse_num(const char *line, unsigned long *p) +{ + char *ptr; + + if (!isdigit(*line)) + return 0; + *p = strtoul(line, &ptr, 10); + return ptr - line; +} + +static int parse_range(const char *line, int len, int offset, const char *expect, + unsigned long *p1, unsigned long *p2) +{ + int digits, ex; + + if (offset < 0 || offset >= len) + return -1; + line += offset; + len -= offset; + + digits = parse_num(line, p1); + if (!digits) + return -1; + + offset += digits; + line += digits; + len -= digits; + + *p2 = 1; + if (*line == ',') { + digits = parse_num(line+1, p2); + if (!digits) + return -1; + + offset += digits+1; + line += digits+1; + len -= digits+1; + } + + ex = strlen(expect); + if (ex > len) + return -1; + if (memcmp(line, expect, ex)) + return -1; + + return offset + ex; +} + +/* + * Parse a unified diff fragment header of the + * form "@@ -a,b +c,d @@" + */ +static int parse_fragment_header(char *line, int len, struct fragment *fragment) +{ + int offset; + + if (!len || line[len-1] != '\n') + return -1; + + /* Figure out the number of lines in a fragment */ + offset = parse_range(line, len, 4, " +", &fragment->oldpos, &fragment->oldlines); + offset = parse_range(line, len, offset, " @@", &fragment->newpos, &fragment->newlines); + + return offset; +} + +static int find_header(char *line, unsigned long size, int *hdrsize, struct patch *patch) +{ + unsigned long offset, len; + + patch->is_rename = patch->is_copy = 0; + patch->is_new = patch->is_delete = -1; + patch->old_mode = patch->new_mode = 0; + patch->old_name = patch->new_name = NULL; + for (offset = 0; size > 0; offset += len, size -= len, line += len, linenr++) { + unsigned long nextlen; + + len = linelen(line, size); + if (!len) + break; + + /* Testing this early allows us to take a few shortcuts.. */ + if (len < 6) + continue; + + /* + * Make sure we don't find any unconnected patch fragmants. + * That's a sign that we didn't find a header, and that a + * patch has become corrupted/broken up. + */ + if (!memcmp("@@ -", line, 4)) { + struct fragment dummy; + if (parse_fragment_header(line, len, &dummy) < 0) + continue; + error("patch fragment without header at line %d: %.*s", linenr, (int)len-1, line); + } + + if (size < len + 6) + break; + + /* + * Git patch? It might not have a real patch, just a rename + * or mode change, so we handle that specially + */ + if (!memcmp("diff --git ", line, 11)) { + int git_hdr_len = parse_git_header(line, len, size, patch); + if (git_hdr_len <= len) + continue; + if (!patch->old_name && !patch->new_name) { + if (!patch->def_name) + die("git diff header lacks filename information (line %d)", linenr); + patch->old_name = patch->new_name = patch->def_name; + } + *hdrsize = git_hdr_len; + return offset; + } + + /** --- followed by +++ ? */ + if (memcmp("--- ", line, 4) || memcmp("+++ ", line + len, 4)) + continue; + + /* + * We only accept unified patches, so we want it to + * at least have "@@ -a,b +c,d @@\n", which is 14 chars + * minimum + */ + nextlen = linelen(line + len, size - len); + if (size < nextlen + 14 || memcmp("@@ -", line + len + nextlen, 4)) + continue; + + /* Ok, we'll consider it a patch */ + parse_traditional_patch(line, line+len, patch); + *hdrsize = len + nextlen; + linenr += 2; + return offset; + } + return -1; +} + +/* + * Parse a unified diff. Note that this really needs + * to parse each fragment separately, since the only + * way to know the difference between a "---" that is + * part of a patch, and a "---" that starts the next + * patch is to look at the line counts.. + */ +static int parse_fragment(char *line, unsigned long size, struct patch *patch, struct fragment *fragment) +{ + int added, deleted; + int len = linelen(line, size), offset; + unsigned long oldlines, newlines; + unsigned long leading, trailing; + + offset = parse_fragment_header(line, len, fragment); + if (offset < 0) + return -1; + oldlines = fragment->oldlines; + newlines = fragment->newlines; + leading = 0; + trailing = 0; + + if (patch->is_new < 0) { + patch->is_new = !oldlines; + if (!oldlines) + patch->old_name = NULL; + } + if (patch->is_delete < 0) { + patch->is_delete = !newlines; + if (!newlines) + patch->new_name = NULL; + } + + if (patch->is_new && oldlines) + return error("new file depends on old contents"); + if (patch->is_delete != !newlines) { + if (newlines) + return error("deleted file still has contents"); + fprintf(stderr, "** warning: file %s becomes empty but is not deleted\n", patch->new_name); + } + + /* Parse the thing.. */ + line += len; + size -= len; + linenr++; + added = deleted = 0; + for (offset = len; size > 0; offset += len, size -= len, line += len, linenr++) { + if (!oldlines && !newlines) + break; + len = linelen(line, size); + if (!len || line[len-1] != '\n') + return -1; + switch (*line) { + default: + return -1; + case ' ': + oldlines--; + newlines--; + if (!deleted && !added) + leading++; + trailing++; + break; + case '-': + deleted++; + oldlines--; + trailing = 0; + break; + case '+': + /* + * We know len is at least two, since we have a '+' and + * we checked that the last character was a '\n' above. + * That is, an addition of an empty line would check + * the '+' here. Sneaky... + */ + if ((new_whitespace != nowarn_whitespace) && + isspace(line[len-2])) { + whitespace_error++; + if (squelch_whitespace_errors && + squelch_whitespace_errors < + whitespace_error) + ; + else { + fprintf(stderr, "Adds trailing whitespace.\n%s:%d:%.*s\n", + patch_input_file, + linenr, len-2, line+1); + } + } + added++; + newlines--; + trailing = 0; + break; + + /* We allow "\ No newline at end of file". Depending + * on locale settings when the patch was produced we + * don't know what this line looks like. The only + * thing we do know is that it begins with "\ ". + * Checking for 12 is just for sanity check -- any + * l10n of "\ No newline..." is at least that long. + */ + case '\\': + if (len < 12 || memcmp(line, "\\ ", 2)) + return -1; + break; + } + } + if (oldlines || newlines) + return -1; + fragment->leading = leading; + fragment->trailing = trailing; + + /* If a fragment ends with an incomplete line, we failed to include + * it in the above loop because we hit oldlines == newlines == 0 + * before seeing it. + */ + if (12 < size && !memcmp(line, "\\ ", 2)) + offset += linelen(line, size); + + patch->lines_added += added; + patch->lines_deleted += deleted; + return offset; +} + +static int parse_single_patch(char *line, unsigned long size, struct patch *patch) +{ + unsigned long offset = 0; + struct fragment **fragp = &patch->fragments; + + while (size > 4 && !memcmp(line, "@@ -", 4)) { + struct fragment *fragment; + int len; + + fragment = xcalloc(1, sizeof(*fragment)); + len = parse_fragment(line, size, patch, fragment); + if (len <= 0) + die("corrupt patch at line %d", linenr); + + fragment->patch = line; + fragment->size = len; + + *fragp = fragment; + fragp = &fragment->next; + + offset += len; + line += len; + size -= len; + } + return offset; +} + +static inline int metadata_changes(struct patch *patch) +{ + return patch->is_rename > 0 || + patch->is_copy > 0 || + patch->is_new > 0 || + patch->is_delete || + (patch->old_mode && patch->new_mode && + patch->old_mode != patch->new_mode); +} + +static int parse_binary(char *buffer, unsigned long size, struct patch *patch) +{ + /* We have read "GIT binary patch\n"; what follows is a line + * that says the patch method (currently, either "deflated + * literal" or "deflated delta") and the length of data before + * deflating; a sequence of 'length-byte' followed by base-85 + * encoded data follows. + * + * Each 5-byte sequence of base-85 encodes up to 4 bytes, + * and we would limit the patch line to 66 characters, + * so one line can fit up to 13 groups that would decode + * to 52 bytes max. The length byte 'A'-'Z' corresponds + * to 1-26 bytes, and 'a'-'z' corresponds to 27-52 bytes. + * The end of binary is signalled with an empty line. + */ + int llen, used; + struct fragment *fragment; + char *data = NULL; + + patch->fragments = fragment = xcalloc(1, sizeof(*fragment)); + + /* Grab the type of patch */ + llen = linelen(buffer, size); + used = llen; + linenr++; + + if (!strncmp(buffer, "delta ", 6)) { + patch->is_binary = BINARY_DELTA_DEFLATED; + patch->deflate_origlen = strtoul(buffer + 6, NULL, 10); + } + else if (!strncmp(buffer, "literal ", 8)) { + patch->is_binary = BINARY_LITERAL_DEFLATED; + patch->deflate_origlen = strtoul(buffer + 8, NULL, 10); + } + else + return error("unrecognized binary patch at line %d: %.*s", + linenr-1, llen-1, buffer); + buffer += llen; + while (1) { + int byte_length, max_byte_length, newsize; + llen = linelen(buffer, size); + used += llen; + linenr++; + if (llen == 1) + break; + /* Minimum line is "A00000\n" which is 7-byte long, + * and the line length must be multiple of 5 plus 2. + */ + if ((llen < 7) || (llen-2) % 5) + goto corrupt; + max_byte_length = (llen - 2) / 5 * 4; + byte_length = *buffer; + if ('A' <= byte_length && byte_length <= 'Z') + byte_length = byte_length - 'A' + 1; + else if ('a' <= byte_length && byte_length <= 'z') + byte_length = byte_length - 'a' + 27; + else + goto corrupt; + /* if the input length was not multiple of 4, we would + * have filler at the end but the filler should never + * exceed 3 bytes + */ + if (max_byte_length < byte_length || + byte_length <= max_byte_length - 4) + goto corrupt; + newsize = fragment->size + byte_length; + data = xrealloc(data, newsize); + if (decode_85(data + fragment->size, + buffer + 1, + byte_length)) + goto corrupt; + fragment->size = newsize; + buffer += llen; + size -= llen; + } + fragment->patch = data; + return used; + corrupt: + return error("corrupt binary patch at line %d: %.*s", + linenr-1, llen-1, buffer); +} + +static int parse_chunk(char *buffer, unsigned long size, struct patch *patch) +{ + int hdrsize, patchsize; + int offset = find_header(buffer, size, &hdrsize, patch); + + if (offset < 0) + return offset; + + patchsize = parse_single_patch(buffer + offset + hdrsize, size - offset - hdrsize, patch); + + if (!patchsize) { + static const char *binhdr[] = { + "Binary files ", + "Files ", + NULL, + }; + static const char git_binary[] = "GIT binary patch\n"; + int i; + int hd = hdrsize + offset; + unsigned long llen = linelen(buffer + hd, size - hd); + + if (llen == sizeof(git_binary) - 1 && + !memcmp(git_binary, buffer + hd, llen)) { + int used; + linenr++; + used = parse_binary(buffer + hd + llen, + size - hd - llen, patch); + if (used) + patchsize = used + llen; + else + patchsize = 0; + } + else if (!memcmp(" differ\n", buffer + hd + llen - 8, 8)) { + for (i = 0; binhdr[i]; i++) { + int len = strlen(binhdr[i]); + if (len < size - hd && + !memcmp(binhdr[i], buffer + hd, len)) { + linenr++; + patch->is_binary = 1; + patchsize = llen; + break; + } + } + } + + /* Empty patch cannot be applied if: + * - it is a binary patch and we do not do binary_replace, or + * - text patch without metadata change + */ + if ((apply || check) && + (patch->is_binary + ? !allow_binary_replacement + : !metadata_changes(patch))) + die("patch with only garbage at line %d", linenr); + } + + return offset + hdrsize + patchsize; +} + +static const char pluses[] = "++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++"; +static const char minuses[]= "----------------------------------------------------------------------"; + +static void show_stats(struct patch *patch) +{ + const char *prefix = ""; + char *name = patch->new_name; + char *qname = NULL; + int len, max, add, del, total; + + if (!name) + name = patch->old_name; + + if (0 < (len = quote_c_style(name, NULL, NULL, 0))) { + qname = xmalloc(len + 1); + quote_c_style(name, qname, NULL, 0); + name = qname; + } + + /* + * "scale" the filename + */ + len = strlen(name); + max = max_len; + if (max > 50) + max = 50; + if (len > max) { + char *slash; + prefix = "..."; + max -= 3; + name += len - max; + slash = strchr(name, '/'); + if (slash) + name = slash; + } + len = max; + + /* + * scale the add/delete + */ + max = max_change; + if (max + len > 70) + max = 70 - len; + + add = patch->lines_added; + del = patch->lines_deleted; + total = add + del; + + if (max_change > 0) { + total = (total * max + max_change / 2) / max_change; + add = (add * max + max_change / 2) / max_change; + del = total - add; + } + if (patch->is_binary) + printf(" %s%-*s | Bin\n", prefix, len, name); + else + printf(" %s%-*s |%5d %.*s%.*s\n", prefix, + len, name, patch->lines_added + patch->lines_deleted, + add, pluses, del, minuses); + if (qname) + free(qname); +} + +static int read_old_data(struct stat *st, const char *path, void *buf, unsigned long size) +{ + int fd; + unsigned long got; + + switch (st->st_mode & S_IFMT) { + case S_IFLNK: + return readlink(path, buf, size); + case S_IFREG: + fd = open(path, O_RDONLY); + if (fd < 0) + return error("unable to open %s", path); + got = 0; + for (;;) { + int ret = xread(fd, buf + got, size - got); + if (ret <= 0) + break; + got += ret; + } + close(fd); + return got; + + default: + return -1; + } +} + +static int find_offset(const char *buf, unsigned long size, const char *fragment, unsigned long fragsize, int line, int *lines) +{ + int i; + unsigned long start, backwards, forwards; + + if (fragsize > size) + return -1; + + start = 0; + if (line > 1) { + unsigned long offset = 0; + i = line-1; + while (offset + fragsize <= size) { + if (buf[offset++] == '\n') { + start = offset; + if (!--i) + break; + } + } + } + + /* Exact line number? */ + if (!memcmp(buf + start, fragment, fragsize)) + return start; + + /* + * There's probably some smart way to do this, but I'll leave + * that to the smart and beautiful people. I'm simple and stupid. + */ + backwards = start; + forwards = start; + for (i = 0; ; i++) { + unsigned long try; |