.gitattributes

# Define the workspace as a DataFlex language project.
# Once DataFlex lands in github-linguist (PR #7479) these mappings switch
# from "suppress misdetection" to "count as DataFlex" automatically.
*.src linguist-language=DataFlex
*.pkg linguist-language=DataFlex
*.vw linguist-language=DataFlex
*.dg linguist-language=DataFlex
*.sl linguist-language=DataFlex
*.bp linguist-language=DataFlex
*.wo linguist-language=DataFlex
*.inc linguist-language=DataFlex
*.rv linguist-language=DataFlex
*.dd linguist-language=DataFlex
*.h linguist-language=DataFlex
*.Inc linguist-language=DataFlex

# Workspace config is INI-format data; without this, *.ws is detected as
# "Witcher Script" (a game-modding language).
*.ws linguist-language=INI

# DataFlex source is written by the Studio with CRLF; the repository stores it
# with LF. Pinning that here keeps a Studio save from showing as a whole-file
# diff. This renormalizes nothing: every text blob in this repository was
# already LF when the pin was added, measured byte by byte. The .sws files are
# deliberately left unpinned - Filesystem26.0.sws is CRLF in the repository and
# pinning it would be a real content change.
*.src text eol=crlf
*.pkg text eol=crlf
*.vw  text eol=crlf
*.dg  text eol=crlf
*.sl  text eol=crlf
*.bp  text eol=crlf
*.wo  text eol=crlf
*.inc text eol=crlf
*.rv  text eol=crlf
*.dd  text eol=crlf
*.frm text eol=crlf

# Windows helper scripts must stay CRLF so cmd.exe runs them correctly
*.bat text eol=crlf
*.cmd text eol=crlf

# The rest of the repository's text files had drifted to CRLF on disk while the
# repository held LF, so every one of them reported as wholly modified. Pinning
# them ends that too, and changes no content: these blobs are LF already.
*.cfg      text eol=crlf
*.fd       text eol=crlf
*.xml      text eol=crlf
*.tag      text eol=crlf
*.ws       text eol=crlf
*.md       text eol=crlf
*.manifest text eol=crlf
.gitignore text eol=crlf
LICENSE    text eol=crlf