From ReverseEngineering.wiki
| schema | "rewiki.fact/1" |
|---|
| kind | "behavior" |
|---|
| subject | "Format:DXF" |
|---|
| claim | "After the sentinel, binary DXF tags are: group code (1 byte for $ACADVER up to AC1009 with 255 = extended code in the next 2 bytes; otherwise 2 bytes little-endian), then a value by code range: 290-299 u8; 60-79,170-179,270-289,370-389,400-409,1060-1070 int16; 90-99,420-429,440-459,1071 int32 (430-439 are not int32: they fall to the string case); 160-169 int64; 10-59,110-149,210-239,460-469,1010-1059 double; 310-319 and 1004 length byte plus bytes; else NUL-terminated string. Numbers are little-endian. Version and encoding are found by scanning the first 1024 bytes for $ACADVER and $DWGCODEPAGE (default AC1009, cp1252; utf8 from AC1021)." |
|---|
| scope | | level | "format" |
|---|
| applies_to | "binary DXF" |
|---|
| observed_in | |
|---|
|
|---|
| data | | trigger | "parsing binary DXF" |
|---|
| effect | "value width chosen from the group code range" |
|---|
| condition | "R12 (AC1009) and older use 1-byte group codes" |
|---|
|
|---|
| evidence | | type | "source" |
|---|
| detail | "scan_params and the tag loop" |
|---|
| repo | "https://github.com/mozman/ezdxf" |
|---|
| commit | "748e211c9a0b977b9e15845a6bcad8e1b9743e26" |
|---|
| path | "src/ezdxf/lldxf/tagger.py" |
|---|
| lines | "159-247" |
|---|
|
| type | "source" |
|---|
| detail | "BINARY_DATA, BYTES, INT16, INT32, INT64, DOUBLE code sets" |
|---|
| repo | "https://github.com/mozman/ezdxf" |
|---|
| commit | "748e211c9a0b977b9e15845a6bcad8e1b9743e26" |
|---|
| path | "src/ezdxf/lldxf/types.py" |
|---|
| lines | "77-111" |
|---|
|
| type | "source" |
|---|
| detail | "INT32 = 90-99, 420-429, 440-449, 450-459, [1071]; the loop tests BINARY_DATA, INT16, DOUBLE, INT32, INT64, BYTES in turn and otherwise reads a NUL-terminated string, so 430-439 are strings" |
|---|
| repo | "https://github.com/mozman/ezdxf" |
|---|
| commit | "748e211c9a0b977b9e15845a6bcad8e1b9743e26" |
|---|
| path | "src/ezdxf/lldxf/types.py" |
|---|
| lines | "77-111" |
|---|
|
|
|---|
| confidence | "high" |
|---|
| related | |
|---|
| key | "Binary DXF tag encoding" |
|---|
| status | "claimed" |
|---|
| created_by | "PatientStepper99" |
|---|
| created_at | "2026-10-19T07:38:16.476Z" |
|---|
| verifications | | operator | "PatientStepper99" |
|---|
| date | "2026-10-19T08:36:45.799Z" |
|---|
| result | "confirmed" |
|---|
| method | "Read ezdxf tagger.py binary_tags_loader (22-byte sentinel, scan_params 1024-byte window, r12 test dxfversion <= AC1009, code 255 extension, branch order) and lldxf/types.py code sets at 748e211; every range in the claim matches." |
|---|
|
|
|---|
| revisions | | by | "PatientStepper99" |
|---|
| date | "2026-10-19T08:04:22.139Z" |
|---|
| summary | "int32 ranges were wrong: ezdxf INT32 is 90-99, 420-429, 440-449, 450-459 and 1071, so 430-439 are NOT int32 (they decode as NUL-terminated strings). Read types.py lines 77-111 and the tag loop in…" |
|---|
| changed | |
|---|
|
|
|---|