=================================================================== RCS file: /cvs/mandoc/cgi.c,v retrieving revision 1.23 retrieving revision 1.33 diff -u -p -r1.23 -r1.33 --- mandoc/cgi.c 2011/12/10 10:59:21 1.23 +++ mandoc/cgi.c 2011/12/15 12:18:57 1.33 @@ -1,4 +1,4 @@ -/* $Id: cgi.c,v 1.23 2011/12/10 10:59:21 kristaps Exp $ */ +/* $Id: cgi.c,v 1.33 2011/12/15 12:18:57 kristaps Exp $ */ /* * Copyright (c) 2011 Kristaps Dzonsons * @@ -24,6 +24,7 @@ #include #include #include +#include #include #include #include @@ -54,46 +55,43 @@ enum page { PAGE__MAX }; +struct paths { + char *name; + char *path; +}; + /* * A query as passed to the search function. - * See kval_query() on how this is parsed. */ struct query { const char *arch; /* architecture */ const char *sec; /* manual section */ const char *expr; /* unparsed expression string */ + int manroot; /* manroot index (or -1)*/ int whatis; /* whether whatis mode */ int legacy; /* whether legacy mode */ }; -struct kval { - char *key; - char *val; -}; - struct req { - struct kval *fields; - size_t fieldsz; + struct query q; + struct paths *p; + size_t psz; enum page page; }; static int atou(const char *, unsigned *); -static void catman(const char *); +static void catman(const struct req *, const char *); static int cmp(const void *, const void *); -static void format(const char *); +static void format(const struct req *, const char *); static void html_print(const char *); static void html_putchar(char); -static int kval_decode(char *); -static void kval_free(struct kval *, size_t); -static void kval_parse(struct kval **, size_t *, char *); -static void kval_query(struct query *, - const struct kval *, size_t); -static void pg_index(const struct manpaths *, - const struct req *, char *); -static void pg_search(const struct manpaths *, - const struct req *, char *); -static void pg_show(const struct manpaths *, - const struct req *, char *); +static int http_decode(char *); +static void http_parse(struct req *, char *); +static int pathstop(DIR *); +static void pathgen(DIR *, char *, struct req *); +static void pg_index(const struct req *, char *); +static void pg_search(const struct req *, char *); +static void pg_show(const struct req *, char *); static void resp_bad(void); static void resp_baddb(void); static void resp_error400(void); @@ -105,9 +103,10 @@ static void resp_index(const struct req *); static void resp_search(struct res *, size_t, void *); static void resp_searchform(const struct req *); -static const char *progname; -static const char *cache; -static const char *host; +static const char *progname; /* cgi script name */ +static const char *cache; /* cache directory */ +static const char *css; /* css directory */ +static const char *host; /* hostname */ static const char * const pages[PAGE__MAX] = { "index", /* PAGE_INDEX */ @@ -116,63 +115,6 @@ static const char * const pages[PAGE__MAX] = { }; /* - * Initialise and parse a query structure from input. - * This accomodates for mdocml's man.cgi and also for legacy man.cgi - * input keys ("sektion" and "apropos"). - * Note that legacy mode has some quirks: if apropos legacy mode is - * detected, we unset the section and architecture string. - */ -static void -kval_query(struct query *q, const struct kval *fields, size_t sz) -{ - int i, legacy; - - memset(q, 0, sizeof(struct query)); - q->whatis = 1; - legacy = -1; - - for (i = 0; i < (int)sz; i++) - if (0 == strcmp(fields[i].key, "expr")) - q->expr = fields[i].val; - else if (0 == strcmp(fields[i].key, "query")) - q->expr = fields[i].val; - else if (0 == strcmp(fields[i].key, "sec")) - q->sec = fields[i].val; - else if (0 == strcmp(fields[i].key, "sektion")) - q->sec = fields[i].val; - else if (0 == strcmp(fields[i].key, "arch")) - q->arch = fields[i].val; - else if (0 == strcmp(fields[i].key, "apropos")) - legacy = 0 == strcmp - (fields[i].val, "0"); - else if (0 == strcmp(fields[i].key, "op")) - q->whatis = 0 == strcasecmp - (fields[i].val, "whatis"); - - /* Test for old man.cgi compatibility mode. */ - - if (legacy == 0) { - q->whatis = 0; - q->legacy = 1; - } else if (legacy > 0) { - q->legacy = 1; - q->whatis = 1; - } - - /* - * Section "0" means no section when in legacy mode. - * For some man.cgi scripts, "default" arch is none. - */ - - if (q->legacy && NULL != q->sec) - if (0 == strcmp(q->sec, "0")) - q->sec = NULL; - if (q->legacy && NULL != q->arch) - if (0 == strcmp(q->arch, "default")) - q->arch = NULL; -} - -/* * This is just OpenBSD's strtol(3) suggestion. * I use it instead of strtonum(3) for portability's sake. */ @@ -236,31 +178,24 @@ html_print(const char *p) html_putchar(*p++); } -static void -kval_free(struct kval *p, size_t sz) -{ - int i; - - for (i = 0; i < (int)sz; i++) { - free(p[i].key); - free(p[i].val); - } - free(p); -} - /* * Parse out key-value pairs from an HTTP request variable. * This can be either a cookie or a POST/GET string, although man.cgi * uses only GET for simplicity. */ static void -kval_parse(struct kval **kv, size_t *kvsz, char *p) +http_parse(struct req *req, char *p) { - char *key, *val; - size_t sz, cur; + char *key, *val, *manroot; + size_t sz; + int i, legacy; - cur = 0; + memset(&req->q, 0, sizeof(struct query)); + req->q.whatis = 1; + legacy = -1; + manroot = NULL; + while (p && '\0' != *p) { while (' ' == *p) p++; @@ -294,21 +229,59 @@ kval_parse(struct kval **kv, size_t *kvsz, char *p) /* Just abort handling. */ - if ( ! kval_decode(key)) - return; - if ( ! kval_decode(val)) - return; + if ( ! http_decode(key)) + break; + if ( ! http_decode(val)) + break; - if (*kvsz + 1 >= cur) { - cur++; - *kv = mandoc_realloc - (*kv, cur * sizeof(struct kval)); - } + if (0 == strcmp(key, "expr")) + req->q.expr = val; + else if (0 == strcmp(key, "query")) + req->q.expr = val; + else if (0 == strcmp(key, "sec")) + req->q.sec = val; + else if (0 == strcmp(key, "sektion")) + req->q.sec = val; + else if (0 == strcmp(key, "arch")) + req->q.arch = val; + else if (0 == strcmp(key, "manpath")) + manroot = val; + else if (0 == strcmp(key, "apropos")) + legacy = 0 == strcmp(val, "0"); + else if (0 == strcmp(key, "op")) + req->q.whatis = 0 == strcasecmp(val, "whatis"); + } - (*kv)[(int)*kvsz].key = mandoc_strdup(key); - (*kv)[(int)*kvsz].val = mandoc_strdup(val); - (*kvsz)++; + /* Test for old man.cgi compatibility mode. */ + + if (legacy == 0) { + req->q.whatis = 0; + req->q.legacy = 1; + } else if (legacy > 0) { + req->q.legacy = 1; + req->q.whatis = 1; } + + /* + * Section "0" means no section when in legacy mode. + * For some man.cgi scripts, "default" arch is none. + */ + + if (req->q.legacy && NULL != req->q.sec) + if (0 == strcmp(req->q.sec, "0")) + req->q.sec = NULL; + if (req->q.legacy && NULL != req->q.arch) + if (0 == strcmp(req->q.arch, "default")) + req->q.arch = NULL; + + /* Default to first manroot. */ + + if (NULL != manroot) { + for (i = 0; i < (int)req->psz; i++) + if (0 == strcmp(req->p[i].name, manroot)) + break; + req->q.manroot = i < (int)req->psz ? i : -1; + } } /* @@ -317,7 +290,7 @@ kval_parse(struct kval **kv, size_t *kvsz, char *p) * over the allocated string. */ static int -kval_decode(char *p) +http_decode(char *p) { char hex[3]; int c; @@ -366,19 +339,21 @@ resp_begin_html(int code, const char *msg) resp_begin_http(code, msg); - puts("\n" - "\n" - "\n" - "\n" - "\n" - "System Manpage Reference\n" - "\n" - "\n" - ""); + printf("\n" + "\n" + "\n" + "\n" + "\n" + "\n" + "System Manpage Reference\n" + "\n" + "\n" + "\n", css, css); } static void @@ -392,34 +367,48 @@ resp_end_html(void) static void resp_searchform(const struct req *req) { - struct query q; + int i; - kval_query(&q, req->fields, req->fieldsz); - puts(""); - printf("
\n"); - printf("
\n" + printf("
\n" + "\n" + "
\n" "Search Parameters\n" " or \n" " for manuals satisfying \n" - "q.expr ? req->q.expr : ""); printf("\">, section " "q.sec ? req->q.sec : ""); printf("\">, arch " ".\n" + html_print(req->q.arch ? req->q.arch : ""); + printf("\">"); + if (req->psz > 1) { + puts(", "); + } + puts(".\n" "\n" "
\n" - ""); + "\n" + "
"); puts(""); } @@ -485,17 +474,19 @@ static void resp_search(struct res *r, size_t sz, void *arg) { int i; - struct query q; const struct req *req; + req = (const struct req *)arg; + assert(req->q.manroot >= 0); + if (1 == sz) { /* * If we have just one result, then jump there now * without any delay. */ puts("Status: 303 See Other"); - printf("Location: http://%s%s/show/%u/%u.html\n", - host, progname, + printf("Location: http://%s%s/show/%d/%u/%u.html\n", + host, progname, req->q.manroot, r[0].volume, r[0].rec); puts("Content-Type: text/html; charset=utf-8\n"); return; @@ -504,40 +495,38 @@ resp_search(struct res *r, size_t sz, void *arg) qsort(r, sz, sizeof(struct res), cmp); resp_begin_html(200, NULL); - - req = (const struct req *)arg; resp_searchform(req); - kval_query(&q, req->fields, req->fieldsz); + puts("
"); + if (0 == sz) { printf("

\n" - "No %s results found.", - q.whatis ? "whatis" : "apropos"); - if (q.whatis) { - printf("(Try q.whatis ? "whatis" : "apropos"); + if (req->q.whatis) { + printf("(Try q.expr ? req->q.expr : ""); printf("&sec="); - html_print(q.sec ? q.sec : ""); + html_print(req->q.sec ? req->q.sec : ""); printf("&arch="); - html_print(q.arch ? q.arch : ""); + html_print(req->q.arch ? req->q.arch : ""); puts("\">apropos?)"); } puts("

"); + puts("
"); resp_end_html(); return; } - puts("

\n" - ""); + puts("
"); for (i = 0; i < (int)sz; i++) { printf("\n" ""); } - puts("
\n" - "", r[i].volume, r[i].rec); + "", + progname, req->q.manroot, + r[i].volume, r[i].rec); html_print(r[i].title); putchar('('); html_print(r[i].cat); @@ -553,20 +542,21 @@ resp_search(struct res *r, size_t sz, void *arg) "
"); + puts("\n" + ""); resp_end_html(); } /* ARGSUSED */ static void -pg_index(const struct manpaths *ps, const struct req *req, char *path) +pg_index(const struct req *req, char *path) { resp_index(req); } static void -catman(const char *file) +catman(const struct req *req, const char *file) { FILE *f; size_t len; @@ -579,20 +569,9 @@ catman(const char *file) return; } - resp_begin_http(200, NULL); - puts("\n" - "\n" - "\n" - "\n" - "\n" - "System Manpage Reference\n" - "\n" - "\n" - "\n" + resp_begin_html(200, NULL); + resp_searchform(req); + puts("
\n" "
");
 
 	while (NULL != (p = fgetln(f, &len))) {
@@ -705,6 +684,7 @@ catman(const char *file)
 	}
 
 	puts("
\n" + "
\n" "\n" ""); @@ -712,7 +692,7 @@ catman(const char *file) } static void -format(const char *file) +format(const struct req *req, const char *file) { struct mparse *mp; int fd; @@ -736,64 +716,104 @@ format(const char *file) return; } - snprintf(opts, sizeof(opts), "style=/man.css," + snprintf(opts, sizeof(opts), "fragment," "man=%s/search.html?sec=%%S&expr=%%N," /*"includes=/cgi-bin/man.cgi/usr/include/%%I"*/, progname); mparse_result(mp, &mdoc, &man); + if (NULL == man && NULL == mdoc) { + resp_baddb(); + mparse_free(mp); + return; + } + + resp_begin_html(200, NULL); + resp_searchform(req); + vp = html_alloc(opts); - if (NULL != mdoc) { - resp_begin_http(200, NULL); + if (NULL != mdoc) html_mdoc(vp, mdoc); - } else if (NULL != man) { - resp_begin_http(200, NULL); + else html_man(vp, man); - } else - resp_baddb(); + puts("\n" + ""); + html_free(vp); mparse_free(mp); } static void -pg_show(const struct manpaths *ps, const struct req *req, char *path) +pg_show(const struct req *req, char *path) { + struct manpaths ps; char *sub; char file[MAXPATHLEN]; const char *fn, *cp; int rc; - unsigned int vol, rec; + unsigned int vol, rec, mr; DB *idx; DBT key, val; - if (NULL == path) { + idx = NULL; + + /* Parse out mroot, volume, and record from the path. */ + + if (NULL == path || NULL == (sub = strchr(path, '/'))) { resp_error400(); return; - } else if (NULL == (sub = strrchr(path, '/'))) { + } + *sub++ = '\0'; + if ( ! atou(path, &mr)) { resp_error400(); return; - } else - *sub++ = '\0'; - - if ( ! (atou(path, &vol) && atou(sub, &rec))) { + } + path = sub; + if (NULL == (sub = strchr(path, '/'))) { resp_error400(); return; - } else if (vol >= (unsigned int)ps->sz) { + } + *sub++ = '\0'; + if ( ! atou(path, &vol) || ! atou(sub, &rec)) { resp_error400(); return; + } else if (mr >= (unsigned int)req->psz) { + resp_error400(); + return; } - strlcpy(file, ps->paths[vol], MAXPATHLEN); + /* + * Begin by chdir()ing into the manroot. + * This way we can pick up the database files, which are + * relative to the manpath root. + */ + + if (-1 == chdir(req->p[(int)mr].path)) { + perror(req->p[(int)mr].path); + resp_baddb(); + return; + } + + memset(&ps, 0, sizeof(struct manpaths)); + manpath_manconf(&ps, "etc/catman.conf"); + + if (vol >= (unsigned int)ps.sz) { + resp_error400(); + goto out; + } + + strlcpy(file, ps.paths[vol], MAXPATHLEN); strlcat(file, "/mandoc.index", MAXPATHLEN); /* Open the index recno(3) database. */ idx = dbopen(file, O_RDONLY, 0, DB_RECNO, NULL); if (NULL == idx) { + perror(file); resp_baddb(); - return; + goto out; } key.data = &rec; @@ -813,42 +833,60 @@ pg_show(const struct manpaths *ps, const struct req *r else if (NULL == memchr(fn, '\0', val.size - (fn - cp))) resp_baddb(); else { - strlcpy(file, cache, MAXPATHLEN); - strlcat(file, "/", MAXPATHLEN); - strlcat(file, fn, MAXPATHLEN); if (0 == strcmp(cp, "cat")) - catman(file); + catman(req, fn + 1); else - format(file); + format(req, fn + 1); } out: - (*idx->close)(idx); + if (idx) + (*idx->close)(idx); + manpath_free(&ps); } static void -pg_search(const struct manpaths *ps, const struct req *req, char *path) +pg_search(const struct req *req, char *path) { size_t tt; + struct manpaths ps; int i, sz, rc; const char *ep, *start; char **cp; struct opts opt; struct expr *expr; - struct query q; - kval_query(&q, req->fields, req->fieldsz); + if (req->q.manroot < 0 || 0 == req->psz) { + resp_search(NULL, 0, (void *)req); + return; + } + memset(&opt, 0, sizeof(struct opts)); - ep = q.expr; - opt.arch = q.arch; - opt.cat = q.sec; + ep = req->q.expr; + opt.arch = req->q.arch; + opt.cat = req->q.sec; rc = -1; sz = 0; cp = NULL; /* - * Poor man's tokenisation. - * Just break apart by spaces. + * Begin by chdir()ing into the root of the manpath. + * This way we can pick up the database files, which are + * relative to the manpath root. + */ + + assert(req->q.manroot < (int)req->psz); + if (-1 == (chdir(req->p[req->q.manroot].path))) { + perror(req->p[req->q.manroot].path); + resp_search(NULL, 0, (void *)req); + return; + } + + memset(&ps, 0, sizeof(struct manpaths)); + manpath_manconf(&ps, "etc/catman.conf"); + + /* + * Poor man's tokenisation: just break apart by spaces. * Yes, this is half-ass. But it works for now. */ @@ -872,12 +910,12 @@ pg_search(const struct manpaths *ps, const struct req * The resp_search() function is called with the results. */ - expr = q.whatis ? termcomp(sz, cp, &tt) : - exprcomp(sz, cp, &tt); + expr = req->q.whatis ? + termcomp(sz, cp, &tt) : exprcomp(sz, cp, &tt); if (NULL != expr) rc = apropos_search - (ps->sz, ps->paths, &opt, + (ps.sz, ps.paths, &opt, expr, tt, (void *)req, resp_search); /* ...unless errors occured. */ @@ -892,41 +930,65 @@ pg_search(const struct manpaths *ps, const struct req free(cp); exprfree(expr); + manpath_free(&ps); } int main(void) { int i; + char buf[MAXPATHLEN]; + DIR *cwd; struct req req; char *p, *path, *subpath; - struct manpaths paths; - /* HTTP init: read and parse the query string. */ + /* Scan our run-time environment. */ - progname = getenv("SCRIPT_NAME"); - if (NULL == progname) + if (NULL == (cache = getenv("CACHE_DIR"))) + cache = "/cache/man.cgi"; + + if (NULL == (progname = getenv("SCRIPT_NAME"))) progname = ""; - cache = getenv("CACHE_DIR"); - if (NULL == cache) - cache = "/cache/man.cgi"; + if (NULL == (css = getenv("CSS_DIR"))) + css = ""; + if (NULL == (host = getenv("HTTP_HOST"))) + host = "localhost"; + + /* + * First we change directory into the cache directory so that + * subsequent scanning for manpath directories is rooted + * relative to the same position. + */ + if (-1 == chdir(cache)) { + perror(cache); resp_bad(); return(EXIT_FAILURE); - } + } else if (NULL == (cwd = opendir(cache))) { + perror(cache); + resp_bad(); + return(EXIT_FAILURE); + } - host = getenv("HTTP_HOST"); - if (NULL == host) - host = "localhost"; - memset(&req, 0, sizeof(struct req)); + strlcpy(buf, ".", MAXPATHLEN); + pathgen(cwd, buf, &req); + closedir(cwd); + + /* Next parse out the query string. */ + if (NULL != (p = getenv("QUERY_STRING"))) - kval_parse(&req.fields, &req.fieldsz, p); + http_parse(&req, p); - /* Resolve leading subpath component. */ + /* + * Now juggle paths to extract information. + * We want to extract our filetype (the file suffix), the + * initial path component, then the trailing component(s). + * Start with leading subpath component. + */ subpath = path = NULL; req.page = PAGE__MAX; @@ -957,31 +1019,29 @@ main(void) break; } - /* Initialise MANPATH. */ - - memset(&paths, 0, sizeof(struct manpaths)); - manpath_manconf("etc/catman.conf", &paths); - /* Route pages. */ switch (req.page) { case (PAGE_INDEX): - pg_index(&paths, &req, subpath); + pg_index(&req, subpath); break; case (PAGE_SEARCH): - pg_search(&paths, &req, subpath); + pg_search(&req, subpath); break; case (PAGE_SHOW): - pg_show(&paths, &req, subpath); + pg_show(&req, subpath); break; default: resp_error404(path); break; } - manpath_free(&paths); - kval_free(req.fields, req.fieldsz); + for (i = 0; i < (int)req.psz; i++) { + free(req.p[i].path); + free(req.p[i].name); + } + free(req.p); return(EXIT_SUCCESS); } @@ -993,3 +1053,115 @@ cmp(const void *p1, const void *p2) ((const struct res *)p2)->title)); } +/* + * Check to see if an "etc" path consists of a catman.conf file. If it + * does, that means that the path contains a tree created by catman(8) + * and should be used for indexing. + */ +static int +pathstop(DIR *dir) +{ + struct dirent *d; + + while (NULL != (d = readdir(dir))) + if (DT_REG == d->d_type) + if (0 == strcmp(d->d_name, "catman.conf")) + return(1); + + return(0); +} + +/* + * Scan for indexable paths. + * This adds all paths with "etc/catman.conf" to the buffer. + */ +static void +pathgen(DIR *dir, char *path, struct req *req) +{ + struct dirent *d; + char *cp; + DIR *cd; + int rc; + size_t sz, ssz; + + sz = strlcat(path, "/", MAXPATHLEN); + if (sz >= MAXPATHLEN) { + fprintf(stderr, "%s: Path too long", path); + return; + } + + /* + * First, scan for the "etc" directory. + * If it's found, then see if it should cause us to stop. This + * happens when a catman.conf is found in the directory. + */ + + rc = 0; + while (0 == rc && NULL != (d = readdir(dir))) { + if (DT_DIR != d->d_type || strcmp(d->d_name, "etc")) + continue; + + path[(int)sz] = '\0'; + ssz = strlcat(path, d->d_name, MAXPATHLEN); + + if (ssz >= MAXPATHLEN) { + fprintf(stderr, "%s: Path too long", path); + return; + } else if (NULL == (cd = opendir(path))) { + perror(path); + return; + } + + rc = pathstop(cd); + closedir(cd); + } + + if (rc > 0) { + /* This also strips the trailing slash. */ + path[(int)--sz] = '\0'; + req->p = mandoc_realloc + (req->p, + (req->psz + 1) * sizeof(struct paths)); + /* + * Strip out the leading "./" unless we're just a ".", + * in which case use an empty string as our name. + */ + req->p[(int)req->psz].path = mandoc_strdup(path); + req->p[(int)req->psz].name = + cp = mandoc_strdup(path + (1 == sz ? 1 : 2)); + req->psz++; + /* + * The name is just the path with all the slashes taken + * out of it. Simple but effective. + */ + for ( ; '\0' != *cp; cp++) + if ('/' == *cp) + *cp = ' '; + return; + } + + /* + * If no etc/catman.conf was found, recursively enter child + * directory and continue scanning. + */ + + rewinddir(dir); + while (NULL != (d = readdir(dir))) { + if (DT_DIR != d->d_type || '.' == d->d_name[0]) + continue; + + path[(int)sz] = '\0'; + ssz = strlcat(path, d->d_name, MAXPATHLEN); + + if (ssz >= MAXPATHLEN) { + fprintf(stderr, "%s: Path too long", path); + return; + } else if (NULL == (cd = opendir(path))) { + perror(path); + return; + } + + pathgen(cd, path, req); + closedir(cd); + } +}