From 3e91311ae750af9bf2e3517b1e701288ac3066b9 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Tue, 24 May 2005 01:07:31 -0700 Subject: [PATCH] Add "cvs2git" program to convert a CVS archive into a GIT archive It's very hacky, and it needs lots of work, but it seems to have converted Peter's "syslinux" archive successfully. Whether the end result is correct or not is to be seen. Tons of work still to do: do name conversion properly, and do tags etc. And testing. Lots of testing. --- Makefile | 2 +- cvs2git.c | 255 ++++++++++++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 256 insertions(+), 1 deletion(-) create mode 100644 cvs2git.c diff --git a/Makefile b/Makefile index b5be0c94ef..99986829f8 100644 --- a/Makefile +++ b/Makefile @@ -2,7 +2,7 @@ CC=gcc CFLAGS=-Wall -O2 HOME=$(shell echo $$HOME) -PROGRAMS=mailsplit mailinfo stripspace +PROGRAMS=mailsplit mailinfo stripspace cvs2git SCRIPTS=dotest applypatch all: $(PROGRAMS) diff --git a/cvs2git.c b/cvs2git.c new file mode 100644 index 0000000000..c18f904f59 --- /dev/null +++ b/cvs2git.c @@ -0,0 +1,255 @@ +/* + * cvs2git + * + * Copyright (C) Linus Torvalds 2005 + */ + +#include +#include +#include +#include +#include + +static int verbose = 0; + +/* + * This is a really stupid program that takes cvsps output, and + * generates a a long _shell_script_ that will create the GIT archive + * from it. + * + * You've been warned. I told you it was stupid. + * + * NOTE NOTE NOTE! In order to do branches correctly, this needs + * the fixed cvsps that has the "Ancestor branch" tag output. + * Hopefully David Mansfield will update his distribution soon + * enough (he's the one who wrote the patch, so at least we don't + * have to figt maintainer issues ;) + */ +enum state { + Header, + Log, + Members +}; + +static char *rcsdir; + +static char date[100]; +static char author[100]; +static char branch[100]; +static char ancestor[100]; +static char tag[100]; +static char log[32768]; +static int loglen = 0; +static int initial_commit = 1; + +static void lookup_author(char *n, char **name, char **email) +{ + /* + * FIXME!!! I'm lazy and stupid. + * + * This could be something like + * + * printf("lookup_author '%s'\n", n); + * *name = "$author_name"; + * *email = "$author_email"; + * + * and that would allow the script to do its own + * lookups at run-time. + */ + *name = n; + *email = n; +} + +static void prepare_commit(void) +{ + char *author_name, *author_email; + char *src_branch; + + lookup_author(author, &author_name, &author_email); + + printf("export GIT_COMMITTER_NAME=%s\n", author_name); + printf("export GIT_COMMITTER_EMAIL=%s\n", author_email); + + printf("export GIT_AUTHOR_NAME=%s\n", author_name); + printf("export GIT_AUTHOR_EMAIL=%s\n", author_email); + + printf("export GIT_AUTHOR_DATE='%s'\n", date); + + if (initial_commit) + return; + + src_branch = *ancestor ? ancestor : branch; + if (!strcmp(src_branch, "HEAD")) + src_branch = "master"; + printf("ln -sf refs/heads/'%s' .git/HEAD\n", src_branch); + printf("git-read-tree -m HEAD\n"); + printf("git-checkout-cache -f -u -a\n"); +} + +static void commit(void) +{ + const char *cmit_parent = initial_commit ? "" : "-p HEAD"; + + printf("tree=$(git-write-tree)\n"); + printf("cat > .cmitmsg < .git/HEAD\n"); + + *date = 0; + *author = 0; + *branch = 0; + *ancestor = 0; + *tag = 0; + loglen = 0; + + initial_commit = 0; +} + +static void get_rcs_name(char *rcspathname, char *name, char *dir) +{ + sprintf(rcspathname, "%s/%s,v", rcsdir, name); + if (!access(rcspathname, R_OK)) + return; + + sprintf(rcspathname, "%s/Attic/%s,v", rcsdir, name); + if (!access(rcspathname, R_OK)) + return; + + if (dir) { + sprintf(rcspathname, "%s/%.*s/Attic/%s,v", rcsdir, (int)(dir - name), name, dir+1); + if (!access(rcspathname, R_OK)) + return; + } + fprintf(stderr, "Unable to find RCS file for %s\n", name); + exit(1); +} + +static void update_file(char *line) +{ + static char rcspathname[4096]; + char *name, *version; + char *dir; + + while (isspace(*line)) + line++; + name = line; + line = strchr(line, ':'); + if (!line) + return; + *line++ = 0; + line = strchr(line, '>'); + if (!line) + return; + *line++ = 0; + version = line; + line = strchr(line, '('); + if (line) { /* "(DEAD)" */ + printf("git-update-cache --force-remove '%s'\n", name); + return; + } + + dir = strrchr(name, '/'); + if (dir) + printf("mkdir -p %.*s\n", (int)(dir - name), name); + + get_rcs_name(rcspathname, name, dir); + + printf("co -p -r%s '%s' > '%s'\n", version, rcspathname, name); + printf("git-update-cache --add -- '%s'\n", name); +} + +struct hdrentry { + const char *name; + char *dest; +} hdrs[] = { + { "Date:", date }, + { "Author:", author }, + { "Branch:", branch }, + { "Ancestor branch:", ancestor }, + { "Tag:", tag }, + { "Log:", NULL }, + { NULL, NULL } +}; + +int main(int argc, char **argv) +{ + static char line[1000]; + enum state state = Header; + + rcsdir = getenv("RCSDIR"); + if (!rcsdir) { + fprintf(stderr, "I need an $RCSDIR\n"); + exit(1); + } + + printf("[ -d .git ] && exit 1\n"); + printf("git-init-db\n"); + printf("mkdir -p .git/refs/heads\n"); + printf("mkdir -p .git/refs/tags\n"); + printf("ln -sf refs/heads/master .git/HEAD\n"); + + while (fgets(line, sizeof(line), stdin) != NULL) { + int linelen = strlen(line); + + while (linelen && isspace(line[linelen-1])) + line[--linelen] = 0; + + switch (state) { + struct hdrentry *entry; + + case Header: + if (verbose) + printf("# H: %s\n", line); + for (entry = hdrs ; entry->name ; entry++) { + int len = strlen(entry->name); + char *val; + + if (memcmp(entry->name, line, len)) + continue; + if (!entry->dest) { + state = Log; + break; + } + val = line + len; + linelen -= len; + while (isspace(*val)) { + val++; + linelen--; + } + memcpy(entry->dest, val, linelen+1); + break; + } + continue; + + case Log: + if (verbose) + printf("# L: %s\n", line); + if (!strcmp(line, "Members:")) { + while (loglen && isspace(log[loglen-1])) + log[--loglen] = 0; + prepare_commit(); + state = Members; + continue; + } + + if (loglen + linelen + 5 > sizeof(log)) + continue; + memcpy(log + loglen, line, linelen); + loglen += linelen; + log[loglen++] = '\n'; + continue; + + case Members: + if (verbose) + printf("# M: %s\n", line); + if (!linelen) { + commit(); + state = Header; + continue; + } + update_file(line); + continue; + } + } + return 0; +}