refactor: simplify main (#141)

* simplified main * linter * bug fix * bump rebdhuhn * remove existing dir by default * remove only files * bump rebdhuhn
Hochfrequenz · Jan 29, 2025 · 70742fe · 70742fe
1 parent 90f26ef
commit 70742fe
Show file tree

Hide file tree

Showing 4 changed files with 54 additions and 88 deletions.
diff --git a/.pre-commit-config.yaml b/.pre-commit-config.yaml
@@ -1,18 +1,18 @@
 # to update all repo revisions just run: pre-commit autoupdate
 repos:
   - repo: https://github.com/pre-commit/pre-commit-hooks
-    rev: v4.4.0
+    rev: v5.0.0
     hooks:
       - id: check-yaml
       - id: end-of-file-fixer
       - id: trailing-whitespace
   - repo: https://github.com/psf/black
-    rev: 23.9.1
+    rev: 24.10.0
     hooks:
       - id: black
         language_version: python3
   - repo: https://github.com/pycqa/isort
-    rev: 5.12.0
+    rev: 5.13.2
     hooks:
       - id: isort
         name: isort (python)

diff --git a/pyproject.toml b/pyproject.toml
@@ -19,7 +19,7 @@ classifiers = [
 ]
 dependencies = [
     "ebdamame>=0.4.1",
-    "rebdhuhn>=0.9.0",
+    "rebdhuhn>=0.14.6",
     "cattrs",
     "click",
     "pydantic-settings"

diff --git a/requirements.txt b/requirements.txt
@@ -48,7 +48,7 @@ python-docx==1.1.2
     # via ebdamame
 python-dotenv==1.0.1
     # via pydantic-settings
-rebdhuhn==0.14.2
+rebdhuhn==0.14.6
     # via
     #   ebd-toolchain (pyproject.toml)
     #   ebdamame

diff --git a/src/ebd_toolchain/main.py b/src/ebd_toolchain/main.py
@@ -34,7 +34,7 @@
 from ebdamame.docxtableconverter import DocxTableConverter
 from pydantic import Field
 from pydantic_settings import BaseSettings, SettingsConfigDict
-from rebdhuhn.graph_conversion import convert_empty_table_to_graph, convert_table_to_graph
+from rebdhuhn.graph_conversion import convert_table_to_graph
 from rebdhuhn.graphviz import convert_dot_to_svg_kroki, convert_graph_to_dot
 from rebdhuhn.kroki import DotToSvgConverter, Kroki, KrokiDotBadRequestError, KrokiPlantUmlBadRequestError
 from rebdhuhn.models.ebd_graph import EbdGraph
@@ -82,16 +82,7 @@ def _dump_svg(svg_path: Path, ebd_graph: EbdGraph, converter: DotToSvgConverter)
 
 def _dump_json(json_path: Path, ebd_table: EbdTable | EbdTableMetaData) -> None:
     with open(json_path, "w+", encoding="utf-8") as json_file:
-        if isinstance(ebd_table, EbdTableMetaData):
-            json.dump(
-                cattrs.unstructure(EbdTable(metadata=ebd_table, rows=[])),
-                json_file,
-                ensure_ascii=False,
-                indent=2,
-                sort_keys=True,
-            )
-        else:
-            json.dump(cattrs.unstructure(ebd_table), json_file, ensure_ascii=False, indent=2, sort_keys=True)
+        json.dump(cattrs.unstructure(ebd_table), json_file, ensure_ascii=False, indent=2, sort_keys=True)
 
 
 @click.command()
@@ -130,11 +121,13 @@ def _main(input_path: Path, output_path: Path, export_types: list[Literal["puml"
     settings = Settings()  # type:ignore[call-arg]
     # read settings from environment variable/.env file
     kroki_client = Kroki(kroki_host=f"http://{settings.kroki_host}:{settings.kroki_port}")
-    if output_path.exists():
-        click.secho(f"The output directory '{output_path}' exists already.", fg="yellow")
-    else:
-        output_path.mkdir(parents=True)
-        click.secho(f"Created a new directory at {output_path}", fg="green")
+    if output_path.exists() and output_path.is_dir():
+        click.secho(f"The output directory '{output_path}' exists already. Will remove its content.", fg="yellow")
+        for item in output_path.iterdir():
+            if item.is_file():
+                item.unlink()
+    output_path.mkdir(parents=True, exist_ok=True)
+    click.secho(f"Created a new directory at {output_path}", fg="green")
     all_ebd_keys = get_all_ebd_keys(input_path)
     error_sources: dict[type, list[str]] = {}
 
@@ -153,59 +146,29 @@ def handle_known_error(error: Exception, ebd_key: str) -> None:
             continue
         assert ebd_kapitel is not None
         assert ebd_kapitel.subsection_title is not None
-        if isinstance(docx_tables, EbdNoTableSection):
-            ebd_meta_data = EbdTableMetaData(
-                ebd_code=ebd_key,
-                ebd_name=ebd_kapitel.subsection_title,
-                chapter=ebd_kapitel.chapter_title,  # type:ignore[arg-type]
-                # pylint:disable=line-too-long
-                section=f"{ebd_kapitel.chapter}.{ebd_kapitel.section}.{ebd_kapitel.subsection}: {ebd_kapitel.section_title}",
-                role="N/A",
-                remark=docx_tables.remark,
-            )
-            if "json" in export_types:
-                json_path = output_path / Path(f"{ebd_key}.json")
-                _dump_json(json_path, ebd_meta_data)
-                click.secho(f"💾 Successfully exported '{ebd_key}.json' to {json_path.absolute()}")
-            try:
-                ebd_graph = convert_empty_table_to_graph(ebd_meta_data)
-            except (
-                EbdCrossReferenceNotSupportedError,
-                EndeInWrongColumnError,
-                OutcomeCodeAmbiguousError,
-            ) as known_issue:
-                handle_known_error(known_issue, ebd_key)
-                continue
-            except Exception as unknown_error:  # pylint:disable=broad-except
-                click.secho(f"Error while graphing {ebd_key}: {str(unknown_error)}; Skip!", fg="red")
-                continue
-            try:
-                if "dot" in export_types:
-                    dot_path = output_path / Path(f"{ebd_key}.dot")
-                    _dump_dot(dot_path, ebd_graph)
-                    click.secho(f"💾 Successfully exported '{ebd_key}.dot' to {dot_path.absolute()}")
-                if "svg" in export_types:
-                    svg_path = output_path / Path(f"{ebd_key}.svg")
-                    _dump_svg(svg_path, ebd_graph, kroki_client)
-                    click.secho(f"💾 Successfully exported '{ebd_key}.svg' to {svg_path.absolute()}")
-            except (PathsNotGreaterThanOneError, KrokiDotBadRequestError) as known_issue:
-                handle_known_error(known_issue, ebd_key)
-            except AssertionError as assertion_error:
-                # e.g. AssertionError: If indegree > 1, the number of paths should always be greater than 1 too.
-                click.secho(str(assertion_error), fg="red")
-                # both the SVG and dot path require graphviz to work, hence the common error handling block
-            continue
         try:
-            assert not isinstance(docx_tables, EbdNoTableSection)
-            converter = DocxTableConverter(
-                docx_tables,
-                ebd_key=ebd_key,
-                ebd_name=ebd_kapitel.subsection_title,
-                chapter=ebd_kapitel.chapter_title,  # type:ignore[arg-type]
-                # pylint:disable=line-too-long
-                section=f"{ebd_kapitel.chapter}.{ebd_kapitel.section}.{ebd_kapitel.subsection}: {ebd_kapitel.section_title}",
-            )
-            ebd_table = converter.convert_docx_tables_to_ebd_table()
+            if isinstance(docx_tables, EbdNoTableSection):
+                ebd_meta_data = EbdTableMetaData(
+                    ebd_code=ebd_key,
+                    ebd_name=ebd_kapitel.subsection_title,
+                    chapter=ebd_kapitel.chapter_title,  # type:ignore[arg-type]
+                    # pylint:disable=line-too-long
+                    section=f"{ebd_kapitel.chapter}.{ebd_kapitel.section}.{ebd_kapitel.subsection}: {ebd_kapitel.section_title}",
+                    role="N/A",
+                    remark=docx_tables.remark,
+                )
+                ebd_table = EbdTable(metadata=ebd_meta_data, rows=[])
+
+            else:
+                converter = DocxTableConverter(
+                    docx_tables,
+                    ebd_key=ebd_key,
+                    ebd_name=ebd_kapitel.subsection_title,
+                    chapter=ebd_kapitel.chapter_title,  # type:ignore[arg-type]
+                    # pylint:disable=line-too-long
+                    section=f"{ebd_kapitel.chapter}.{ebd_kapitel.section}.{ebd_kapitel.subsection}: {ebd_kapitel.section_title}",
+                )
+                ebd_table = converter.convert_docx_tables_to_ebd_table()
         except Exception as scraping_error:  # pylint:disable=broad-except
             click.secho(f"Error while scraping {ebd_key}: {str(scraping_error)}; Skip!", fg="red")
             continue
@@ -222,21 +185,24 @@ def handle_known_error(error: Exception, ebd_key: str) -> None:
             click.secho(f"Error while graphing {ebd_key}: {str(unknown_error)}; Skip!", fg="red")
             continue
         if "puml" in export_types:
-            try:
-                puml_path = output_path / Path(f"{ebd_key}.puml")
-                _dump_puml(puml_path, ebd_graph)
-                click.secho(f"💾 Successfully exported '{ebd_key}.puml' to {puml_path.absolute()}")
-            except AssertionError as assertion_error:
-                # https://github.com/Hochfrequenz/rebdhuhn/issues/35
-                click.secho(str(assertion_error), fg="red")
-            except (
-                NotExactlyTwoOutgoingEdgesError,
-                GraphTooComplexForPlantumlError,
-                KrokiPlantUmlBadRequestError,
-            ) as known_issue:
-                handle_known_error(known_issue, ebd_key)
-            except Exception as general_error:  # pylint:disable=broad-exception-caught
-                click.secho(f"Error while exporting {ebd_key} as UML: {str(general_error)}; Skip!", fg="yellow")
+            if not any(ebd_table.rows):
+                click.secho(f"EBD {ebd_key} has no ebd table; Skip puml creation!", fg="yellow")
+            else:
+                try:
+                    puml_path = output_path / Path(f"{ebd_key}.puml")
+                    _dump_puml(puml_path, ebd_graph)
+                    click.secho(f"💾 Successfully exported '{ebd_key}.puml' to {puml_path.absolute()}")
+                except AssertionError as assertion_error:
+                    # https://github.com/Hochfrequenz/rebdhuhn/issues/35
+                    click.secho(str(assertion_error), fg="red")
+                except (
+                    NotExactlyTwoOutgoingEdgesError,
+                    GraphTooComplexForPlantumlError,
+                    KrokiPlantUmlBadRequestError,
+                ) as known_issue:
+                    handle_known_error(known_issue, ebd_key)
+                except Exception as general_error:  # pylint:disable=broad-exception-caught
+                    click.secho(f"Error while exporting {ebd_key} as UML: {str(general_error)}; Skip!", fg="yellow")
 
         try:
             if "dot" in export_types: