Ok, ok, still in the "Variables" section, R2010, reading the string chain, handles came after reading 1 bit and 2 bytes more.
But this is commented out in the last LibreDWG code for R2007, "refactoring" branch, any idea why?. See file "header.c", line 45. 2014/1/17 Till Heuschmann <[email protected]> > Yes, 2 bytes. > > > Am 17.01.2014 um 18:49 schrieb Felipe Castro <[email protected]>: > > That´s correct Till, it works also for 2 other samples I have here. I have > tested for the string chain, now I will go for the handles. > > Other little question, in a DWG file does Unicode mean UCS-2 (2 bytes for > each character)? > > > Thanks! > > > > 2014/1/16 Till Heuschmann <[email protected]> > >> Hi Felipe, >> >> a section from a R2010 file (at least some of my sample files) is build >> like this: >> >> - sentinel (16 bytes) >> - data_size (4 bytes) >> - if maintenance_release_version is > 3 than an unknown long follows (4 >> bytes) >> - bit_size (4 bytes) >> >> With the bit_size the location of the string stream could be calculated >> like in a R2007 file with one exception: >> in the 2007 string_stream_init() function the bitpos variable is >> incremented by 0x8F >> in the 2010 string_stream_init() function the bitpos variable must be >> incremented by 0xAF >> >> The maintenance_release_version variable mentioned earlier is the >> variable from the file header. >> >> >> Am 16.01.2014 um 18:42 schrieb Felipe Castro: >> >> Well, I didn´t found any of those streams after the integer and float >> variables of the header. >> >> While reading the variables, the only handle to be read directly is the >> HANDSEED, the others should be read from some handle stream. The same with >> strings, they should be fetched from some stream in other place. >> >> But in R2010, there is no bitsize to get the string stream after the >> variables (as with R2007). >> >> Also, after reading the variables, a bit 0 is found, no way to get a >> string stream from there. >> >> So, I have searched for strings and handles after reading the variables, >> and did not succeed anyway. >> >> >> >> >> 2014/1/10 Felipe Castro <[email protected]> >> >>> Till, only after 2 months I'm having the time to get back looking at the >>> code here, thanks for your suggestion, I'll take a look at those string >>> offsets. >>> >>> >>> 2013/11/11 Till Heuschmann <[email protected]> >>> >>>> I would suggest that the string and handle chains are handled like in >>>> R2007. Have a look at the r2007 branch and if you are lucky not much has to >>>> be changed. >>>> >>>> Am 11.11.2013 um 10:46 schrieb Felipe Castro: >>>> >>>> >>>> I was trying to decode R2010, but I have an issue, where to get the >>>> string chain for the header variables section. Anyone can help on this with >>>> more information? >>>> >>>> >>>> >> >> >
