sub: add filter text utils, use from filter-regex (no-op)

Add two stand-alone function to help with the text-extraction task
which ass filters need. Makes it easier to add new filters without
cargo-culting this functionality.

Currently, on malformed event (which shouldn't happen), a warning is
printed when a filter tries to extract the text, so if few filters
are enabled, we'll get multiple warnings (like before) - not critical.

The regex filter now uses these utils, the SDH filter not yet.

sub/filter_regex.c

@@ -44,15 +44,7 @@ static bool rf_init(struct sd_filter *ft)
     if (!p->num_regexes)
         return false;
 
-    char *headers = ft->event_format;
-    while (headers && headers[0]) {
-        p->offset += 1;
-        headers = strchr(headers, ',');
-        if (headers)
-            headers += 1;
-    }
-    p->offset -= 1; // removes Start/End, adds ReadOrder
-
+    p->offset = sd_ass_fmt_offset(ft->event_format);
     return true;
 }
 
@@ -68,20 +60,9 @@ static struct demux_packet *rf_filter(struct sd_filter *ft,
                                       struct demux_packet *pkt)
 {
     struct priv *p = ft->priv;
-    char *line = bstrto0(NULL, (bstr){(char *)pkt->buffer, pkt->len});
+    char *text = bstrto0(NULL, sd_ass_pkt_text(ft, pkt, p->offset));
     bool drop = false;
 
-    char *text = line;
-    for (int n = 0; n < p->offset - 1; n++) {
-        text = strchr(text, ',');
-        if (!text) {
-            MP_WARN(ft, "Malformed event: '%s'\n", line);
-            text = line; // shouldn't happen; random fallback
-            break;
-        }
-        text = text + 1;
-    }
-
     for (int n = 0; n < p->num_regexes; n++) {
         int err = regexec(&p->regexes[n], text, 0, NULL, 0);
         if (err == 0) {
@@ -94,7 +75,7 @@ static struct demux_packet *rf_filter(struct sd_filter *ft,
         }
     }
 
-    talloc_free(line);
+    talloc_free(text);
     return drop ? NULL : pkt;
 }
 

sub/sd.h

@@ -89,4 +89,15 @@ struct sd_filter_functions {
 extern const struct sd_filter_functions sd_filter_sdh;
 extern const struct sd_filter_functions sd_filter_regex;
 
+
+// convenience utils for filters with ass codec
+
+// num commas to skip at an ass-event before the "Text" field (always last)
+// (doesn't change, can be retrieved once on filter init)
+int sd_ass_fmt_offset(const char *event_format);
+
+// the event (pkt->buffer) "Text" content according to the calculated offset.
+// on malformed event: warns and returns (bstr){NULL,0}
+bstr sd_ass_pkt_text(struct sd_filter *ft, struct demux_packet *pkt, int offset);
+
 #endif

sub/sd_ass.c

@@ -945,3 +945,28 @@ static void mangle_colors(struct sd *sd, struct sub_bitmaps *parts)
         sb->libass.color = MP_ASS_RGBA(rgb[0], rgb[1], rgb[2], a);
     }
 }
+
+int sd_ass_fmt_offset(const char *evt_fmt)
+{
+    // "Text" is always last (as it's arbitrary content in buf), e.g. format:
+    // "Layer, Start, End, Style, Name, MarginL, MarginR, MarginV, Effect, Text"
+    int n = 0;
+    while (evt_fmt && (evt_fmt = strchr(evt_fmt, ',')))
+         evt_fmt++, n++;
+    return n-1;  // buffer is without the format's Start/End, with ReadOrder
+}
+
+bstr sd_ass_pkt_text(struct sd_filter *ft, struct demux_packet *pkt, int offset)
+{
+    // e.g. pkt->buffer ("4" is ReadOrder): "4,0,Default,,0,0,0,,fifth line"
+    bstr txt = {(char *)pkt->buffer, pkt->len}, t0 = txt;
+    while (offset-- > 0) {
+        int n = bstrchr(txt, ',');
+        if (n < 0) {  // shouldn't happen
+            MP_WARN(ft, "Malformed event '%.*s'\n", BSTR_P(t0));
+            return (bstr){NULL, 0};
+        }
+        txt = bstr_cut(txt, n+1);
+    }
+    return txt;
+}