##// END OF EJS Templates
subrepo: make -S work again on Windows for incoming/outgoing to remote repos...
subrepo: make -S work again on Windows for incoming/outgoing to remote repos While it worked fine for the top level repo, the remote path for the subrepo got mangled to something like "https://server/prefix\repo\subrepo", which I've seen result in both a 400 and a 404, depending on the server. We need to `normpath` at least the `subpath` because of "http://user:***@localhost:$HGPORT/main/../sub" in `test-subrepo-relative-path.t`. Keep the `os.path` flavor for handling filesystem based remote repos, since this string is also displayed. This is one case where the automatic substitution of '\' for '/' and rematching done by the test runner is unfortunate- I don't see how to write a test to catch this. Differential Revision: https://phab.mercurial-scm.org/D11971

File last commit:

r35977:9724f549 default
r49400:f98d4d0a stable
Show More
hgclient.h
30 lines | 866 B | text/x-c | CLexer
/*
* A command server client that uses Unix domain socket
*
* Copyright (c) 2011 Yuya Nishihara <yuya@tcha.org>
*
* This software may be used and distributed according to the terms of the
* GNU General Public License version 2 or any later version.
*/
#ifndef HGCLIENT_H_
#define HGCLIENT_H_
#include <sys/types.h>
struct hgclient_tag_;
typedef struct hgclient_tag_ hgclient_t;
hgclient_t *hgc_open(const char *sockname);
void hgc_close(hgclient_t *hgc);
pid_t hgc_peerpgid(const hgclient_t *hgc);
pid_t hgc_peerpid(const hgclient_t *hgc);
const char **hgc_validate(hgclient_t *hgc, const char *const args[],
size_t argsize);
int hgc_runcommand(hgclient_t *hgc, const char *const args[], size_t argsize);
void hgc_attachio(hgclient_t *hgc);
void hgc_setenv(hgclient_t *hgc, const char *const envp[]);
#endif /* HGCLIENT_H_ */