OpenSCAD merge

Problem

OpenSCAD is a great parametric-design CAD file.

It allows including libraries and other files into the main code file.

Which is fine for ordinary operation. Less fine for code distribution, especially for one-off use.

A single file without dependencies is then called for.

Can also allow freezing the functioning finished file and making it immune to library changes/updates.

Solution

mergescad.py was written.

The file is loaded into an array. Each time an include or use statement is encountered, the specified file is recursively loaded.

Only the variables/functions/modules used are extracted. Dead weight is left out.

Python was chosen for this level of complexity, as the startup time penalty is not worth the necessary cost of not spending too much time writing it all in C and bash is WAY too weak.

Usage

options

usage: mergescad.py [-h] [-q] [-o] [-F] [-O OUTFILE] [-L [LIBDIR]] [-t] [-T]
                    [-I] [-a] [-A] [-d]
                    file.scad

merges together OpenSCAD source files

positional arguments:
  file.scad             input filename, OpenSCAD

optional arguments:
  -h, --help            show this help message and exit
  -q, --quiet           print fewer details
  -o, --tofile          output to filename.merged.scad instead to stdout
  -F, --nofiledet       do not include date/size/name of source files to
                        comments
  -O OUTFILE, --outfile OUTFILE
                        output to specified filename instead to stdout
  -L [LIBDIR], --libdir [LIBDIR]
                        add library directory (multiple possible)
  -t, --tracefile       show directories searched for includes
  -T, --treefile        show directories searched for includes, tree form
  -I, --treefileinc     show includes in tree form, forces -T
  -a, --abspath         show absolute paths, otherwise relative to input file
  -A, --listarr         list internal array (debug)
  -d, --dryrun          dry run, errors output only

output

• -o - instead of default stdout, autogenerate filename.merged.ext from filename.ext and output there
• -O <filename> - like -o but do not autogenerate

paths

• $OPENSCADPATH env variable is employed
• -L allows additional library paths, relative to input file if not absolute

analysis

• -d - dry run, no output of merged file, stderr errors only
• -q - quieter run, do not show included files and output file to stderr

• -t - trace the files as they are open/searched, show path attempts
• -T - tree, show the included files hierarchy tree-wise, useful for third-party libraries-heavy code (libdotscad!)
• -I - tree, also show the include/use statements themselves

• -a - output absolute paths (default is relative to input file), useful when confused
• -F - do not show file details in code comment (name, size, modification time), enabled by default for easier versioning

debug

• -A - show the internal array into which the files are read

caution

The files aren't parsed in any artful way. Identifiers are extracted willy nilly from the code, after comments are removed. Then matching identifiers that look like modules or variables or functions are taken from the included file. Local variables/identifiers are handled as global (TODO...). False positives may happen.

dependencies

• os
• sys
• re
• argparse
• pathlib
• datetime

Download

• mergescad.py

examples

• test-voronoi.scad - random test file, using shadlib.scad and libdotscad
• test-voronoi.merged.scad - merged file without dependencies

TODO

• more commands
• better error handling
• more lege artis parsing of the scad code
• as-is, there may be false positives in identifiers included from the files
• test on Windows