Hi,

-I added few changes to pdf-token-reader.c regarding comments token. Otherwise, 
the main changes are affecting the token-reader unit test.
Left to be done are dictionaries, arrays and streams tokens, then other public 
functions defined in pdf-token-reader.h. 

-About streams, in pdf-token-reader.c we allow the keyword stream to be 
followed by white-space and comments, but I can't find that in the 
specification (PDF Reference , version 1.7) which says page 61: " The keyword 
stream that follows the stream dictionary should be followed by an end-of-line 
marker ...". So why don't we limit the character following stream to an eol. 
Besides, if we allow comments, we should not consider the eol of a comment as a 
eol indicating the beginning of stream data.

-In the unit test pedf-token-read.c, in pdf_token_read_eos is written: /*TODO: 
verify stream position */ what should I do exactly ? 


Best regards, Pierre



      

Attachment: pdf-token-read-patch
Description: Binary data

Reply via email to