Prepare to have a parametrisable path separator character. Currently

the value of 'pathsep' never changes, but the code should now be
ready to cope when it does.

git-svn-id: svn://svn.tartarus.org/sgt/agedu@8241 cda61777-01e9-0310-a592-d414129be87e

diff --git a/TODO b/TODO
index 9b93ca2..1a8a7eb 100644 (file)
--- a/TODO
+++ b/TODO
@@ -56,8 +56,8 @@ Before it's non-embarrassingly releasable:
        * what do we do elsewhere about _GNU_SOURCE?
 
  - prepare a little in advance for a potential future Windows port:
-    + store the separator character in the index file when writing
-      it, and be prepared to cope on reading if it isn't a slash
+    + store the path separator character in the index file when
+      writing it, and load it back in when reading
     + store literal byte sizes in all the size fields, instead of
       Unixoid 512-byte sectors
 
@@ -109,9 +109,16 @@ Future directions:
 
  - http://msdn.microsoft.com/en-us/library/ms724290.aspx suggest
    modern Windowses support atime-equivalents, so a Windows port is
-   possible in principle. Would need to modify the current structure
-   a lot, to abstract away (at least) memory-mapping of files,
-   details of disk scan procedure, networking for httpd. Unclear
-   what the right UI would be on Windows, too; command-line exactly
-   as now might be considered just a _little_ unfriendly. Or perhaps
-   not.
+   possible in principle.
+    + For a full Windows port, would need to modify the current
+      structure a lot, to abstract away (at least) memory-mapping of
+      files, details of disk scan procedure, networking for httpd.
+      Unclear what the right UI would be on Windows, too;
+      command-line exactly as now might be considered just a
+      _little_ unfriendly. Or perhaps not.
+    + Alternatively, a much easier approach would be to write a
+      Windows version of just the --scan-dump mode, which does a
+      filesystem scan via the Windows API and generates a valid
+      agedu dump file on standard output. Then one would simply feed
+      that over the network connection of one's choice to the rest
+      of agedu running on Unix as usual.

diff --git a/agedu.c b/agedu.c
index c88fdf5..d72c73d 100644 (file)
--- a/agedu.c
+++ b/agedu.c
@@ -31,6 +31,16 @@
 
 #define lenof(x) (sizeof((x))/sizeof(*(x)))
 
+/*
+ * Path separator. This global variable affects the behaviour of
+ * various parts of the code when they need to deal with path
+ * separators. The path separator appropriate to a particular data
+ * set is encoded in the index file storing that data set; data
+ * sets generated on Unix will of course have the default '/', but
+ * foreign data sets are conceivable and must be handled correctly.
+ */
+char pathsep = '/';
+
 void fatal(const char *fmt, ...)
 {
     va_list ap;
@@ -86,7 +96,7 @@ static int gotdata(void *vctx, const char *pathname, const struct stat64 *st)
      * Filter based on wildcards.
      */
     include = 1;
-    filename = strrchr(pathname, '/');
+    filename = strrchr(pathname, pathsep);
     if (!filename)
        filename = pathname;
     else
@@ -864,7 +874,7 @@ int main(int argc, char **argv)
         * Trim trailing slash, just in case.
         */
        pathlen = strlen(querydir);
-       if (pathlen > 0 && querydir[pathlen-1] == '/')
+       if (pathlen > 0 && querydir[pathlen-1] == pathsep)
            querydir[--pathlen] = '\0';
 
        text_query(mappedfile, querydir, textcutoff, 1);
@@ -895,7 +905,7 @@ int main(int argc, char **argv)
         * Trim trailing slash, just in case.
         */
        pathlen = strlen(querydir);
-       if (pathlen > 0 && querydir[pathlen-1] == '/')
+       if (pathlen > 0 && querydir[pathlen-1] == pathsep)
            querydir[--pathlen] = '\0';
 
        xi = trie_before(mappedfile, querydir);

diff --git a/html.c b/html.c
index 47e4372..1daabe6 100644 (file)
--- a/html.c
+++ b/html.c
@@ -20,6 +20,8 @@
 
 #define MAXCOLOUR 511
 
+extern char pathsep;
+
 struct html {
     char *buf;
     size_t buflen, bufsize;
@@ -442,7 +444,7 @@ char *html_query(const void *t, unsigned long index,
      */
     htprintf(ctx, "<p align=center>\n<code>");
     q = path;
-    for (p = strchr(path, '/'); p; p = strchr(p+1, '/')) {
+    for (p = strchr(path, pathsep); p; p = strchr(p+1, pathsep)) {
        int doing_href = 0;
        /*
         * See if this path prefix exists in the trie. If so,
@@ -456,12 +458,12 @@ char *html_query(const void *t, unsigned long index,
            htprintf(ctx, "<a href=\"%s\">", href);
            doing_href = 1;
        }
-       *p = '/';
+       *p = pathsep;
        htescape(ctx, q, p - q, 1);
        q = p + 1;
        if (doing_href)
            htprintf(ctx, "</a>");
-       htprintf(ctx, "/");
+       htescape(ctx, q, p - q, 1);
     }
     htescape(ctx, q, strlen(q), 1);
     htprintf(ctx, "</code>\n");

diff --git a/trie.c b/trie.c
index f0b21df..7061c74 100644 (file)
--- a/trie.c
+++ b/trie.c
@@ -16,6 +16,8 @@
 
 #define alignof(typ) ( offsetof(struct { char c; typ t; }, t) )
 
+extern char pathsep;
+
 /*
  * Compare functions for pathnames. Returns the relative order of
  * the names, like strcmp; also passes back the offset of the
@@ -23,9 +25,9 @@
  */
 static int trieccmp(unsigned char a, unsigned char b)
 {
-    a = (a == '\0' ? '\0' : a == '/' ? '\1' : a+1);
-    b = (b == '\0' ? '\0' : b == '/' ? '\1' : b+1);
-    return a - b;
+    a = (a == '\0' ? '\0' : a == pathsep ? '\1' : a+1);
+    b = (b == '\0' ? '\0' : b == pathsep ? '\1' : b+1);
+    return (int)a - (int)b;
 }
 
 static int triencmp(const char *a, size_t alen,