[cfsmp3]

In raising this pull request, I confirm the following (please check boxes):

Reason for this PR:

• This PR adds new functionality.
• This PR fixes a bug that I have personally experienced or that a real user has reported and for which a sample exists.
• This PR is porting code from C to Rust.

Sanity check:

• I have read and understood the contributors guide.
• I have checked that another pull request for this purpose does not exist.
• If the PR adds new functionality, I've added it to the changelog. If it's just a bug fix, I have NOT added it to the changelog.
• I am NOT adding new C code unless it's to fix an existing, reproducible bug.

Repro instructions:

ccextractor 4b117b4d66e81f393a67a07358d143beeb4f384dc9d1c640761f7eeca9f707e7.ts --xmltv 1 (a DVB recording from our sample set) segfaults (exit 139). The 75 EIT packets (PID 0x12) from its first 2 MB, 14,100 bytes, reproduce it on their own; under ASan: heap-buffer-overflow, READ of size 4294967295 in memcpy from EPG_decode_extended_event_descriptor (ts_tables_epg.c:993).

---

EPG_decode_extended_event_descriptor() strips a leading character-table byte from a continuation descriptor (descriptor_number > 0) when the first text byte is below 0x20. When text_length is 0, that byte isn't text but whatever follows the descriptor, and text_length-- wraps the uint32_t to 0xFFFFFFFF, so the following memcpy reads 4 GB. An empty text field is legal — this broadcaster puts all extended event content in the item fields (the crashing call: descriptor 3, text_length 0, next byte 0x09). This only strips the byte when there is text to strip.

Verified:

• The sample and the 14 KB reproducer: segfault / ASan heap-buffer-overflow before; with this change both complete, ASan clean, and the XMLTV file is written (656 programmes).
• 29 local TS samples, default and --xmltv 1: output and exit codes identical to master, except the crashing sample.