glitchlings 0.4.1__tar.gz → 0.4.3__tar.gz

{glitchlings-0.4.1 → glitchlings-0.4.3}/MANIFEST.in

@@ -1,5 +1,5 @@
 recursive-include rust *.rs *.toml *.lock *.tsv
-recursive-include src/glitchlings/zoo *.tsv
+recursive-include src/glitchlings/zoo *.tsv *.json
 recursive-include src/glitchlings/lexicon/data *.json
 include src/glitchlings/config.toml
 prune rust/target

{glitchlings-0.4.1 → glitchlings-0.4.3}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: glitchlings
-Version: 0.4.1
+Version: 0.4.3
 Summary: Monsters for your language games.
 Author: osoleve
 License:                                  Apache License
@@ -226,19 +226,51 @@ License-File: LICENSE
 Requires-Dist: confusable-homoglyphs>=3.3.1
 Requires-Dist: tomli>=2.0.1; python_version < "3.11"
 Requires-Dist: pyyaml>=6.0.0
+Provides-Extra: all
+Requires-Dist: black>=24.4.0; extra == "all"
+Requires-Dist: hypothesis>=6.140.0; extra == "all"
+Requires-Dist: interrogate>=1.5.0; extra == "all"
+Requires-Dist: jellyfish>=1.2.0; extra == "all"
+Requires-Dist: isort>=5.13.0; extra == "all"
+Requires-Dist: mkdocs>=1.6.0; extra == "all"
+Requires-Dist: mkdocs-material>=9.5.0; extra == "all"
+Requires-Dist: mkdocstrings[python]>=0.24.0; extra == "all"
+Requires-Dist: mkdocstrings-python>=1.10.0; extra == "all"
+Requires-Dist: mypy>=1.8.0; extra == "all"
+Requires-Dist: numpy<=2.0,>=1.24; extra == "all"
+Requires-Dist: pre-commit>=3.8.0; extra == "all"
+Requires-Dist: pytest>=8.0.0; extra == "all"
+Requires-Dist: ruff>=0.6.0; extra == "all"
+Requires-Dist: verifiers>=0.1.3.post0; extra == "all"
 Provides-Extra: hf
 Requires-Dist: datasets>=4.0.0; extra == "hf"
+Provides-Extra: lightning
+Requires-Dist: pytorch_lightning>=2.0.0; extra == "lightning"
 Provides-Extra: vectors
 Requires-Dist: numpy<=2.0,>=1.24; extra == "vectors"
 Requires-Dist: spacy>=3.7.2; extra == "vectors"
 Requires-Dist: gensim>=4.3.2; extra == "vectors"
+Provides-Extra: st
+Requires-Dist: sentence-transformers>=3.0.0; extra == "st"
 Provides-Extra: prime
 Requires-Dist: verifiers>=0.1.3.post0; extra == "prime"
 Requires-Dist: jellyfish>=1.2.0; extra == "prime"
+Provides-Extra: torch
+Requires-Dist: torch>=2.0.0; extra == "torch"
 Provides-Extra: dev
 Requires-Dist: pytest>=8.0.0; extra == "dev"
 Requires-Dist: hypothesis>=6.140.0; extra == "dev"
 Requires-Dist: numpy<=2.0,>=1.24; extra == "dev"
+Requires-Dist: mkdocs>=1.6.0; extra == "dev"
+Requires-Dist: mkdocstrings[python]>=0.24.0; extra == "dev"
+Requires-Dist: mkdocs-material>=9.5.0; extra == "dev"
+Requires-Dist: mkdocstrings-python>=1.10.0; extra == "dev"
+Requires-Dist: interrogate>=1.5.0; extra == "dev"
+Requires-Dist: black>=24.4.0; extra == "dev"
+Requires-Dist: isort>=5.13.0; extra == "dev"
+Requires-Dist: ruff>=0.6.0; extra == "dev"
+Requires-Dist: mypy>=1.8.0; extra == "dev"
+Requires-Dist: pre-commit>=3.8.0; extra == "dev"
 Dynamic: license-file
 
 #
@@ -297,7 +329,7 @@ print(gaggle(SAMPLE_TEXT))
 > Onҽ m‎ھ‎rning, wһen Gregor Samƽa woke from trouble𝐝 𝑑reams, he found himself transformed in his bed into a horrible vermin‎٠‎ He l   lay on his armour-like back, and if he lifted his head a little he could see his brown belly, slightlh domed and divided by arches ino stiff sections. The bedding was adly able to cover it and and seemed ready to slide off any  moment. His many legxs, pitifully thin compared with the size of the the rest of him, waved about helplessly ashe looked looked.
 
 Consult the [Glitchlings Usage Guide](docs/index.md)
-for end-to-end instructions spanning the Python API, CLI, HuggingFace and Prime Intellect
+for end-to-end instructions spanning the Python API, CLI, HuggingFace, PyTorch, and Prime Intellect
 integrations, and the autodetected Rust pipeline (enabled whenever the extension is present).
 
 ## Motivation
@@ -338,10 +370,67 @@ They're horrible little gremlins, but they're not _unreasonable_.
 
 Keyboard warriors can challenge them directly via the `glitchlings` command:
 
+<!-- BEGIN: CLI_USAGE -->
 ```bash
 # Discover which glitchlings are currently on the loose.
 glitchlings --list
+```
+
+```text
+   Typogre — scope: Character, order: early
+Apostrofae — scope: Character, order: normal
+     Mim1c — scope: Character, order: last
+  Jargoyle — scope: Word, order: normal
+     Adjax — scope: Word, order: normal
+   Reduple — scope: Word, order: normal
+  Rushmore — scope: Word, order: normal
+  Redactyl — scope: Word, order: normal
+Scannequin — scope: Character, order: late
+    Zeedub — scope: Character, order: last
+```
+
+```bash
+# Review the full CLI contract.
+glitchlings --help
+```
+
+```text
+usage: glitchlings [-h] [-g SPEC] [-s SEED] [-f FILE] [--sample] [--diff]
+                   [--list] [-c CONFIG]
+                   [text]
+
+Summon glitchlings to corrupt text. Provide input text as an argument, via
+--file, or pipe it on stdin.
+
+positional arguments:
+  text                  Text to corrupt. If omitted, stdin is used or --sample
+                        provides fallback text.
+
+options:
+  -h, --help            show this help message and exit
+  -g SPEC, --glitchling SPEC
+                        Glitchling to apply, optionally with parameters like
+                        Typogre(rate=0.05). Repeat for multiples; defaults to
+                        all built-ins.
+  -s SEED, --seed SEED  Seed controlling deterministic corruption order
+                        (default: 151).
+  -f FILE, --file FILE  Read input text from a file instead of the command
+                        line argument.
+  --sample              Use the included SAMPLE_TEXT when no other input is
+                        provided.
+  --diff                Show a unified diff between the original and corrupted
+                        text.
+  --list                List available glitchlings and exit.
+  -c CONFIG, --config CONFIG
+                        Load glitchlings from a YAML configuration file.
+```
+<!-- END: CLI_USAGE -->
+
+Run `python docs/build_cli_reference.py` whenever you tweak the CLI so the README stays in sync with the actual output. The script executes the commands above and replaces the block between the markers automatically.
 
+Prefer inline tweaks? You can still configure glitchlings directly in the shell:
+
+```bash
 # Run Typogre against the contents of a file and inspect the diff.
 glitchlings -g typogre --file documents/report.txt --diff
 
@@ -355,8 +444,6 @@ echo "Beware LLM-written flavor-text" | glitchlings -g mim1c
 glitchlings --config experiments/chaos.yaml "Let slips the glitchlings of war"
 ```
 
-Use `--help` for a complete breakdown of available options, including support for parameterised glitchlings via `-g "Name(arg=value, ...)"` to mirror the Python API.
-
 Attack configurations live in plain YAML files so you can version-control experiments without touching code:
 
 ```yaml
@@ -394,6 +481,16 @@ _What a nice word, would be a shame if something happened to it._
 > - `keyboard (str)`: Keyboard layout key-neighbor map to use (default: "CURATOR_QWERTY"; also accepts "QWERTY", "DVORAK", "COLEMAK", and "AZERTY").
 > - `seed (int)`: The random seed for reproducibility (default: 151).
 
+### Apostrofae
+
+_It looks like you're trying to paste some text. Can I help?_
+
+> _**Paperclip Manager.**_ Apostrofae scans for balanced runs of straight quotes, apostrophes, and backticks before replacing them with randomly sampled smart-quote pairs from a curated lookup table. The swap happens in-place so contractions and unpaired glyphs remain untouched.
+>
+> Args
+>
+> - `seed (int)`: Optional seed controlling the deterministic smart-quote sampling (default: 151).
+
 ### Mim1c
 
 _Wait, was that...?_

{glitchlings-0.4.1 → glitchlings-0.4.3}/README.md

@@ -54,7 +54,7 @@ print(gaggle(SAMPLE_TEXT))
 > Onҽ m‎ھ‎rning, wһen Gregor Samƽa woke from trouble𝐝 𝑑reams, he found himself transformed in his bed into a horrible vermin‎٠‎ He l   lay on his armour-like back, and if he lifted his head a little he could see his brown belly, slightlh domed and divided by arches ino stiff sections. The bedding was adly able to cover it and and seemed ready to slide off any  moment. His many legxs, pitifully thin compared with the size of the the rest of him, waved about helplessly ashe looked looked.
 
 Consult the [Glitchlings Usage Guide](docs/index.md)
-for end-to-end instructions spanning the Python API, CLI, HuggingFace and Prime Intellect
+for end-to-end instructions spanning the Python API, CLI, HuggingFace, PyTorch, and Prime Intellect
 integrations, and the autodetected Rust pipeline (enabled whenever the extension is present).
 
 ## Motivation
@@ -95,10 +95,67 @@ They're horrible little gremlins, but they're not _unreasonable_.
 
 Keyboard warriors can challenge them directly via the `glitchlings` command:
 
+<!-- BEGIN: CLI_USAGE -->
 ```bash
 # Discover which glitchlings are currently on the loose.
 glitchlings --list
+```
+
+```text
+   Typogre — scope: Character, order: early
+Apostrofae — scope: Character, order: normal
+     Mim1c — scope: Character, order: last
+  Jargoyle — scope: Word, order: normal
+     Adjax — scope: Word, order: normal
+   Reduple — scope: Word, order: normal
+  Rushmore — scope: Word, order: normal
+  Redactyl — scope: Word, order: normal
+Scannequin — scope: Character, order: late
+    Zeedub — scope: Character, order: last
+```
+
+```bash
+# Review the full CLI contract.
+glitchlings --help
+```
+
+```text
+usage: glitchlings [-h] [-g SPEC] [-s SEED] [-f FILE] [--sample] [--diff]
+                   [--list] [-c CONFIG]
+                   [text]
+
+Summon glitchlings to corrupt text. Provide input text as an argument, via
+--file, or pipe it on stdin.
+
+positional arguments:
+  text                  Text to corrupt. If omitted, stdin is used or --sample
+                        provides fallback text.
+
+options:
+  -h, --help            show this help message and exit
+  -g SPEC, --glitchling SPEC
+                        Glitchling to apply, optionally with parameters like
+                        Typogre(rate=0.05). Repeat for multiples; defaults to
+                        all built-ins.
+  -s SEED, --seed SEED  Seed controlling deterministic corruption order
+                        (default: 151).
+  -f FILE, --file FILE  Read input text from a file instead of the command
+                        line argument.
+  --sample              Use the included SAMPLE_TEXT when no other input is
+                        provided.
+  --diff                Show a unified diff between the original and corrupted
+                        text.
+  --list                List available glitchlings and exit.
+  -c CONFIG, --config CONFIG
+                        Load glitchlings from a YAML configuration file.
+```
+<!-- END: CLI_USAGE -->
+
+Run `python docs/build_cli_reference.py` whenever you tweak the CLI so the README stays in sync with the actual output. The script executes the commands above and replaces the block between the markers automatically.
 
+Prefer inline tweaks? You can still configure glitchlings directly in the shell:
+
+```bash
 # Run Typogre against the contents of a file and inspect the diff.
 glitchlings -g typogre --file documents/report.txt --diff
 
@@ -112,8 +169,6 @@ echo "Beware LLM-written flavor-text" | glitchlings -g mim1c
 glitchlings --config experiments/chaos.yaml "Let slips the glitchlings of war"
 ```
 
-Use `--help` for a complete breakdown of available options, including support for parameterised glitchlings via `-g "Name(arg=value, ...)"` to mirror the Python API.
-
 Attack configurations live in plain YAML files so you can version-control experiments without touching code:
 
 ```yaml
@@ -151,6 +206,16 @@ _What a nice word, would be a shame if something happened to it._
 > - `keyboard (str)`: Keyboard layout key-neighbor map to use (default: "CURATOR_QWERTY"; also accepts "QWERTY", "DVORAK", "COLEMAK", and "AZERTY").
 > - `seed (int)`: The random seed for reproducibility (default: 151).
 
+### Apostrofae
+
+_It looks like you're trying to paste some text. Can I help?_
+
+> _**Paperclip Manager.**_ Apostrofae scans for balanced runs of straight quotes, apostrophes, and backticks before replacing them with randomly sampled smart-quote pairs from a curated lookup table. The swap happens in-place so contractions and unpaired glyphs remain untouched.
+>
+> Args
+>
+> - `seed (int)`: Optional seed controlling the deterministic smart-quote sampling (default: 151).
+
 ### Mim1c
 
 _Wait, was that...?_

glitchlings-0.4.3/pyproject.toml

@@ -0,0 +1,186 @@
+[project]
+name = "glitchlings"
+version = "0.4.3"
+description = "Monsters for your language games."
+readme = "README.md"
+requires-python = ">=3.10"
+
+dependencies = [
+    "confusable-homoglyphs>=3.3.1",
+    "tomli>=2.0.1; python_version < '3.11'",
+    "pyyaml>=6.0.0",
+]
+
+authors = [
+    { name = "osoleve" }
+]
+
+keywords = [
+    "nlp", 
+    "adversarial augmentation",
+    "text augmentation",
+    "data augmentation",
+    "domain randomization"
+]
+
+classifiers = [
+    "Development Status :: 3 - Alpha",
+    "Intended Audience :: Developers",
+    "Programming Language :: Python",
+    "Programming Language :: Python :: 3",
+    "Programming Language :: Python :: 3.10",
+    "Programming Language :: Python :: 3.11",
+    "Programming Language :: Python :: 3.12",
+    "Programming Language :: Rust",
+    "Topic :: Scientific/Engineering :: Artificial Intelligence",
+    "Topic :: Software Development :: Testing",
+]
+
+[project.license]
+file = "LICENSE"
+
+[project.urls]
+Homepage = "https://github.com/osoleve/glitchlings"
+Repository = "https://github.com/osoleve/glitchlings.git"
+Issues = "https://github.com/osoleve/glitchlings/issues"
+Changelog = "https://github.com/osoleve/glitchlings/releases"
+
+[project.scripts]
+glitchlings = "glitchlings.main:main"
+
+[project.optional-dependencies]
+all = [
+    "black>=24.4.0",
+    "hypothesis>=6.140.0",
+    "interrogate>=1.5.0",
+    "jellyfish>=1.2.0",
+    "isort>=5.13.0",
+    "mkdocs>=1.6.0",
+    "mkdocs-material>=9.5.0",
+    "mkdocstrings[python]>=0.24.0",
+    "mkdocstrings-python>=1.10.0",
+    "mypy>=1.8.0",
+    "numpy>=1.24,<=2.0",
+    "pre-commit>=3.8.0",
+    "pytest>=8.0.0",
+    "ruff>=0.6.0",
+    "verifiers>=0.1.3.post0",
+]
+hf = ["datasets>=4.0.0"]
+lightning = ["pytorch_lightning>=2.0.0"]
+vectors = ["numpy>=1.24,<=2.0", "spacy>=3.7.2", "gensim>=4.3.2"]
+st = ["sentence-transformers>=3.0.0"]
+prime = ["verifiers>=0.1.3.post0", "jellyfish>=1.2.0"]
+torch = ["torch>=2.0.0"]
+dev = [
+    "pytest>=8.0.0",
+    "hypothesis>=6.140.0",
+    "numpy>=1.24,<=2.0",
+    "mkdocs>=1.6.0",
+    "mkdocstrings[python]>=0.24.0",
+    "mkdocs-material>=9.5.0",
+    "mkdocstrings-python>=1.10.0",
+    "interrogate>=1.5.0",
+    "black>=24.4.0",
+    "isort>=5.13.0",
+    "ruff>=0.6.0",
+    "mypy>=1.8.0",
+    "pre-commit>=3.8.0",
+]
+
+[build-system]
+requires = ["setuptools>=69", "wheel", "setuptools-rust>=1.8.0"]
+build-backend = "setuptools.build_meta"
+
+[tool.setuptools]
+package-dir = {"" = "src"}
+include-package-data = true
+
+[tool.setuptools.package-data]
+"glitchlings" = ["config.toml"]
+"glitchlings.lexicon" = ["data/*.json"]
+"glitchlings.zoo" = ["ocr_confusions.tsv"]
+"glitchlings.zoo.assets" = ["apostrofae_pairs.json"]
+
+[tool.setuptools.packages.find]
+where = ["src"]
+
+[[tool.setuptools-rust.ext-modules]]
+target = "glitchlings._zoo_rust"
+path = "rust/zoo/Cargo.toml"
+binding = "PyO3"
+debug = false
+
+
+[tool.pytest.ini_options]
+pythonpath = [
+    "src",
+]
+
+[tool.interrogate]
+config = true
+fail-under = 80
+ignore-init-module = true
+ignore-module = true
+ignore-nested-functions = true
+ignore-private = true
+ignore-semiprivate = true
+ignore-magic = true
+ignore-property-decorators = false
+color = true
+quiet = false
+exclude = [
+    "tests",
+    "docs",
+    "rust",
+    "benchmarks",
+]
+
+[tool.black]
+line-length = 100
+target-version = ["py310"]
+
+[tool.isort]
+profile = "black"
+line_length = 100
+
+[tool.ruff]
+target-version = "py310"
+line-length = 100
+
+[tool.ruff.lint]
+select = ["E", "F", "I"]
+
+[tool.mypy]
+python_version = "3.10"
+follow_imports = "skip"
+ignore_missing_imports = true
+enable_error_code = ["ignore-without-code"]
+
+[[tool.mypy.overrides]]
+module = [
+    "glitchlings.util.adapters",
+    "glitchlings.dlc._shared",
+    "glitchlings.dlc.huggingface",
+    "glitchlings.dlc.prime",
+]
+strict = true
+
+[[tool.mypy.overrides]]
+module = [
+    "glitchlings.compat",
+    "glitchlings.config",
+    "glitchlings.lexicon",
+    "glitchlings.lexicon.*",
+]
+strict = true
+
+[[tool.mypy.overrides]]
+module = [
+    "glitchlings.main",
+    "glitchlings.__main__",
+    "glitchlings.__init__",
+    "glitchlings.zoo",
+    "glitchlings.zoo.*",
+]
+strict = true

{glitchlings-0.4.1 → glitchlings-0.4.3}/rust/Cargo.lock

@@ -90,6 +90,12 @@ version = "2.0.6"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "f4c7245a08504955605670dbf141fceab975f15ca21570696aebe9d2e71576bd"
 
+[[package]]
+name = "itoa"
+version = "1.0.15"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "4a5f13b858c8d314ee3e8f639011f7ccefe71f97f96e50151fb991f267928e2c"
+
 [[package]]
 name = "libc"
 version = "0.2.176"
@@ -275,12 +281,60 @@ version = "0.8.6"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "caf4aa5b0f434c91fe5c7f1ecb6a5ece2130b02ad2a590589dda5146df959001"
 
+[[package]]
+name = "ryu"
+version = "1.0.20"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "28d3b2b1366ec20994f1fd18c3c594f05c5dd4bc44d8bb0c1c632c8d6829481f"
+
 [[package]]
 name = "scopeguard"
 version = "1.2.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "94143f37725109f92c262ed2cf5e59bce7498c01bcc1502d7b9afe439a4e9f49"
 
+[[package]]
+name = "serde"
+version = "1.0.228"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "9a8e94ea7f378bd32cbbd37198a4a91436180c5bb472411e48b5ec2e2124ae9e"
+dependencies = [
+ "serde_core",
+]
+
+[[package]]
+name = "serde_core"
+version = "1.0.228"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "41d385c7d4ca58e59fc732af25c3983b67ac852c1a25000afe1175de458b67ad"
+dependencies = [
+ "serde_derive",
+]
+
+[[package]]
+name = "serde_derive"
+version = "1.0.228"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "d540f220d3187173da220f885ab66608367b6574e925011a9353e4badda91d79"
+dependencies = [
+ "proc-macro2",
+ "quote",
+ "syn",
+]
+
+[[package]]
+name = "serde_json"
+version = "1.0.145"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "402a6f66d8c709116cf22f558eab210f5a50187f702eb4d7e5ef38d9a7f1c79c"
+dependencies = [
+ "itoa",
+ "memchr",
+ "ryu",
+ "serde",
+ "serde_core",
+]
+
 [[package]]
 name = "smallvec"
 version = "1.15.1"
@@ -407,5 +461,6 @@ dependencies = [
  "pyo3",
  "pyo3-build-config",
  "regex",
+ "serde_json",
  "smallvec",
 ]

{glitchlings-0.4.1 → glitchlings-0.4.3}/rust/zoo/Cargo.toml

@@ -14,6 +14,7 @@ regex = { workspace = true }
 once_cell = { workspace = true }
 blake2 = { workspace = true }
 smallvec = "1"
+serde_json = "1"
 
 [package.metadata.maturin]
 module-name = "glitchlings._zoo_rust"

{glitchlings-0.4.1 → glitchlings-0.4.3}/rust/zoo/build.rs

@@ -6,7 +6,9 @@ use std::path::PathBuf;
 use std::process::Command;
 
 fn main() {
-    prepare_confusion_table().expect("failed to stage OCR confusion table for compilation");
+    stage_asset("ocr_confusions.tsv").expect("failed to stage OCR confusion table for compilation");
+    stage_asset("apostrofae_pairs.json")
+        .expect("failed to stage Apostrofae replacement table for compilation");
     pyo3_build_config::add_extension_module_link_args();
 
     // Only perform custom Python linking on non-Linux platforms.
@@ -97,46 +99,60 @@ fn query_python(python: &OsStr, command: &str) -> Option<String> {
     Some(value)
 }
 
-fn prepare_confusion_table() -> io::Result<()> {
+fn stage_asset(asset_name: &str) -> io::Result<()> {
     let manifest_dir = PathBuf::from(env::var("CARGO_MANIFEST_DIR").expect("missing manifest dir"));
     let out_dir = PathBuf::from(env::var("OUT_DIR").expect("missing OUT_DIR"));
 
-    let repo_path = manifest_dir.join("../../src/glitchlings/zoo/ocr_confusions.tsv");
-    let packaged_path = manifest_dir.join("assets/ocr_confusions.tsv");
+    let repo_candidates = [
+        manifest_dir
+            .join("../../src/glitchlings/zoo/assets")
+            .join(asset_name),
+        manifest_dir
+            .join("../../src/glitchlings/zoo")
+            .join(asset_name),
+    ];
+    let packaged_path = manifest_dir.join("assets").join(asset_name);
     println!("cargo:rerun-if-changed={}", packaged_path.display());
 
-    let source_path = if repo_path.exists() {
-        println!("cargo:rerun-if-changed={}", repo_path.display());
-        if packaged_path.exists() {
-            let repo_bytes = fs::read(&repo_path)?;
-            let packaged_bytes = fs::read(&packaged_path)?;
-            if repo_bytes != packaged_bytes {
-                return Err(io::Error::new(
-                    ErrorKind::Other,
-                    format!(
-                        "OCR confusion table at {} is out of sync with {}",
-                        packaged_path.display(),
-                        repo_path.display()
-                    ),
-                ));
+    let mut source_path: Option<PathBuf> = None;
+    for candidate in &repo_candidates {
+        if candidate.exists() {
+            println!("cargo:rerun-if-changed={}", candidate.display());
+            if packaged_path.exists() {
+                let repo_bytes = fs::read(candidate)?;
+                let packaged_bytes = fs::read(&packaged_path)?;
+                if repo_bytes != packaged_bytes {
+                    return Err(io::Error::new(
+                        ErrorKind::Other,
+                        format!(
+                            "asset {} is out of sync with {}",
+                            packaged_path.display(),
+                            candidate.display()
+                        ),
+                    ));
+                }
             }
+            source_path = Some(candidate.clone());
+            break;
         }
-        repo_path
-    } else {
-        if !packaged_path.exists() {
-            return Err(io::Error::new(
-                ErrorKind::NotFound,
-                format!(
-                    "missing OCR confusion table; looked for {} and {}",
-                    repo_path.display(),
-                    packaged_path.display()
-                ),
-            ));
-        }
+    }
+
+    let source_path = if let Some(path) = source_path {
+        path
+    } else if packaged_path.exists() {
         packaged_path
+    } else {
+        return Err(io::Error::new(
+            ErrorKind::NotFound,
+            format!(
+                "missing asset {asset_name}; looked for {} and {}",
+                repo_candidates[0].display(),
+                packaged_path.display()
+            ),
+        ));
     };
 
     fs::create_dir_all(&out_dir)?;
-    fs::copy(&source_path, out_dir.join("ocr_confusions.tsv"))?;
+    fs::copy(&source_path, out_dir.join(asset_name))?;
     Ok(())
 }