Fix install spacy transformer model error(#138)

xxyzz · Jul 24, 2023 · bd77e87 · bd77e87
1 parent cdbaeef
commit bd77e87
Show file tree

Hide file tree

Showing 4 changed files with 14 additions and 5 deletions.
diff --git a/__init__.py b/__init__.py
@@ -2,7 +2,7 @@
 
 from calibre.customize import InterfaceActionBase
 
-VERSION = (3, 29, 4)
+VERSION = (3, 29, 5)
 
 
 class WordDumbDumb(InterfaceActionBase):

diff --git a/data/deps.json b/data/deps.json
@@ -4,5 +4,6 @@
   "thinc-apple-ops": "0.1.3",
   "torch": "2.0.1",
   "rapidfuzz": "3.1.2",
-  "spacy_model": "3.6.0"
+  "spacy_cpu_model": "3.6.0",
+  "spacy_trf_model": "3.6.1"
 }
diff --git a/deps.py b/deps.py
@@ -47,8 +47,11 @@ def install_deps(pkg: str, notif: Any) -> None:
         # Install X-Ray dependencies
         pip_install("rapidfuzz", dep_versions["rapidfuzz"], notif=notif)
 
-        url = f"https://github.com/explosion/spacy-models/releases/download/{pkg}-{dep_versions['spacy_model']}/{pkg}-{dep_versions['spacy_model']}-py3-none-any.whl"
-        pip_install(pkg, dep_versions["spacy_model"], url=url, notif=notif)
+        model_version = dep_versions[
+            "spacy_trf_model" if pkg.endswith("_trf") else "spacy_cpu_model"
+        ]
+        url = f"https://github.com/explosion/spacy-models/releases/download/{pkg}-{model_version}/{pkg}-{model_version}-py3-none-any.whl"
+        pip_install(pkg, model_version, url=url, notif=notif)
         if pkg.endswith("_trf"):
             from .config import prefs
 

diff --git a/parse_job.py b/parse_job.py
@@ -235,6 +235,9 @@ def create_files(
     prefs: Prefs,
     notif: Any,
 ) -> None:
+    """
+    This function runs in system Python subprocess for official(frozen) calibre build.
+    """
     is_epub = not kfx_json and not mobi_codec
     plugin_path = Path(plugin_path_str)
     insert_installed_libs(plugin_path)
@@ -779,7 +782,9 @@ def create_spacy_matcher(
 
     disabled_pipes = list(set(["ner", "parser", "senter"]) & set(nlp.pipe_names))
     pkg_versions = load_plugin_json(plugin_path, "data/deps.json")
-    model_version = pkg_versions["spacy_model"]
+    model_version = model_version = pkg_versions[
+        "spacy_trf_model" if model.endswith("_trf") else "spacy_cpu_model"
+    ]
     phrase_matcher = PhraseMatcher(nlp.vocab, attr="LOWER")
     phrases_doc_path = spacy_doc_path(
         model, model_version, lemma_lang, is_kindle, True, plugin_path, prefs