Refactor for compatibility with macros

Author: timvink

mkdocs_table_reader_plugin/plugin.py

@@ -32,13 +32,47 @@ def on_config(self, config, **kwargs):
         Returns:
             Config
         """
+        self.readers = {reader: READERS[reader].set_config_context(mkdocs_config=config, plugin_config=self.config) for reader in self.config.get('select_readers') if reader in self.config.get('select_readers',[])}
+
         plugins = [p for p in config.get("plugins")]
 
-        for post_load_plugin in ["macros", "markdownextradata"]:
+        # Plugins required before table-reader
+        for post_load_plugin in ["markdownextradata"]:
             if post_load_plugin in plugins:
                 if plugins.index("table-reader") > plugins.index(post_load_plugin):
                     raise ConfigurationError(f"[table-reader]: Incompatible plugin order: Define 'table-reader' before '{post_load_plugin}' in your mkdocs.yml.")
+        
+        # Plugins required after table-reader
+        for post_load_plugin in ["macros"]:
+            if post_load_plugin in plugins:
+                if plugins.index("table-reader") < plugins.index(post_load_plugin):
+                    raise ConfigurationError(f"[table-reader]: Incompatible plugin order: Define 'table-reader' after '{post_load_plugin}' in your mkdocs.yml.")
+
+        if "macros" in config.plugins:
+            config.plugins['macros'].macros.update(self.readers)
+            config.plugins['macros'].variables['macros'].update(self.readers)
+            config.plugins['macros'].env.globals.update(self.readers)
+            self.external_jinja_engine = True
+        else:
+            self.external_jinja_engine = False
+        
+
+    def on_pre_page(self, page, config, **kwargs):
+        """
+        See https://www.mkdocs.org/dev-guide/plugins/#on_pre_page.
+
+        Args:
+            page: mkdocs.nav.Page instance
+            config: global configuration object
 
+        Returns:
+            Page
+        """
+        # store the current page in the plugin config
+        # because the readers have access to the plugin config, they can know where the current page is
+        # this way, they can check that directory too
+        self.config._current_page = page.file.abs_src_path
+        return page
 
     def on_page_markdown(self, markdown, page, config, files, **kwargs):
         """
@@ -60,21 +94,11 @@ def on_page_markdown(self, markdown, page, config, files, **kwargs):
         Returns:
             str: Markdown source text of page as string
         """
-        # Determine the mkdocs directory
-        # We do this during the on_page_markdown() event because other plugins
-        # might have changed the directory.
-        if self.config.get("base_path") == "config_dir":
-            mkdocs_dir = os.path.dirname(os.path.abspath(config["config_file_path"]))
-        if self.config.get("base_path") == "docs_dir":
-            mkdocs_dir = os.path.abspath(config["docs_dir"])
-        
-        # Define directories to search for tables
-        search_directories = [os.path.join(mkdocs_dir, self.config.get("data_path"))]
-        if self.config.get("search_page_directory"):
-            search_directories.append(os.path.dirname(page.file.abs_src_path))
+        if self.external_jinja_engine:
+            return markdown
 
-        for reader in self.config.get('select_readers'):
-            function = READERS[reader]
+        for reader in self.readers:
+            function = self.readers[reader]
             # Regex pattern for tags like {{ read_csv(..) }}
             # match group 0: to extract any leading whitespace 
             # match group 1: to extract the arguments (positional and keywords)
@@ -92,36 +116,11 @@ def on_page_markdown(self, markdown, page, config, files, **kwargs):
 
                 # Safely parse the arguments
                 pd_args, pd_kwargs = parse_argkwarg(result[1])
-
-                # Extract the filepath,
-                # which is the first positional argument
-                # or a named argument when there are no positional arguments
-                if len(pd_args) > 0:
-                    input_file_path = pd_args.pop(0)
-                else:
-                    input_file_path = pd_kwargs.pop("filepath_or_buffer")
-
-                # Validate if file exists
-                search_file_paths = [os.path.join(search_dir, input_file_path) for search_dir in search_directories]
-                valid_file_paths = [p for p in search_file_paths if os.path.exists(p)]
-                if len(valid_file_paths) == 0:
-                    msg = f"[table-reader-plugin]: Cannot find table file '{input_file_path}'. The following directories were searched: {*search_directories,}"
-                    if self.config.get("allow_missing_files"):
-                        logger.warning(msg)
-
-                        # Add message in markdown
-                        updated_tag = fix_indentation(leading_spaces, f"{{{{ Cannot find '{input_file_path}' }}}}")
-
-                        markdown = tag_pattern.sub(updated_tag, markdown, count=1)
-
-                        continue
-                    else:
-                        raise FileNotFoundError(msg)
 
                 # Load the table
                 # note we use the first valid file paths,
                 # where we first search the 'data_path' and then the page's directory.
-                markdown_table = function(valid_file_paths[0], *pd_args, **pd_kwargs)
+                markdown_table = function(*pd_args, **pd_kwargs)
                 markdown_table = fix_indentation(leading_spaces, markdown_table)
 
                 # Insert markdown table

mkdocs_table_reader_plugin/readers.py

@@ -1,19 +1,71 @@
 import pandas as pd
 import yaml
+import os
+from pathlib import Path
+import logging
+
+import functools
 
 from mkdocs_table_reader_plugin.utils import kwargs_in_func, kwargs_not_in_func 
 from mkdocs_table_reader_plugin.markdown import convert_to_md_table
 
-def read_csv(*args, **kwargs):
-    
+logger = logging.getLogger("mkdocs.plugins")
+
+
+class ParseArgs:
+    def __init__(self, func):
+        functools.update_wrapper(self, func)
+        self.func = func
+        self.mkdocs_config = None
+        self.plugin_config = None
+        
+    def set_config_context(self, mkdocs_config, plugin_config):
+        self.mkdocs_config = mkdocs_config
+        self.plugin_config = plugin_config
+        return self
+
+    def __call__(self, *args, **kwargs):
+        assert self.mkdocs_config is not None, "mkdocs_config is not set"
+        assert self.plugin_config is not None, "plugin_config is not set"
+
+        # Extract the filepath,
+        # which is the first positional argument
+        # or a named argument when there are no positional arguments
+        args = list(args)
+        if len(args) > 0:
+            input_file_name = args.pop(0)
+        else:
+            input_file_name = kwargs.pop("filepath_or_buffer")
+
+        possible_file_paths = [
+            Path(os.path.dirname(os.path.abspath(self.mkdocs_config["config_file_path"]))) / Path(self.plugin_config.get("data_path")) / input_file_name,
+            Path(os.path.abspath(self.mkdocs_config["docs_dir"])) / Path(self.plugin_config.get("data_path")) / input_file_name,
+            Path(self.plugin_config._current_page).parent / input_file_name
+        ]
+        valid_file_paths = [path for path in possible_file_paths if path.exists()]
+        if len(valid_file_paths) == 0:
+            msg = f"[table-reader-plugin]: Cannot find table file '{input_file_name}'. The following directories were searched: {*possible_file_paths,}"
+            if self.plugin_config.get("allow_missing_files"):
+                logger.warning(msg)
+                return f"{{{{ Cannot find '{input_file_name}' }}}}"
+            else:
+                raise FileNotFoundError(msg)
+        
+        return self.func(valid_file_paths[0], *args, **kwargs)
+
+
+
+@ParseArgs
+def read_csv(*args, **kwargs) -> str:
     read_kwargs = kwargs_in_func(kwargs, pd.read_csv)
     df = pd.read_csv(*args, **read_kwargs)
 
     markdown_kwargs = kwargs_not_in_func(kwargs, pd.read_csv)
     return convert_to_md_table(df, markdown_kwargs)
 
 
-def read_table(*args, **kwargs):
+@ParseArgs
+def read_table(*args, **kwargs) -> str:
 
     read_kwargs = kwargs_in_func(kwargs, pd.read_table)
     df = pd.read_table(*args, **read_kwargs)
@@ -22,30 +74,35 @@ def read_table(*args, **kwargs):
     return convert_to_md_table(df, markdown_kwargs)
 
 
-def read_fwf(*args, **kwargs):
+@ParseArgs
+def read_fwf(*args, **kwargs) -> str:
     read_kwargs = kwargs_in_func(kwargs, pd.read_fwf)
     df = pd.read_fwf(*args, **read_kwargs)
 
     markdown_kwargs = kwargs_not_in_func(kwargs, pd.read_fwf)
     return convert_to_md_table(df, markdown_kwargs)
 
-def read_json(*args, **kwargs):
+
+@ParseArgs
+def read_json(*args, **kwargs) -> str:
     read_kwargs = kwargs_in_func(kwargs, pd.read_json)
     df = pd.read_json(*args, **read_kwargs)
 
     markdown_kwargs = kwargs_not_in_func(kwargs, pd.read_json)
     return convert_to_md_table(df, markdown_kwargs)
 
 
-def read_excel(*args, **kwargs):
+@ParseArgs
+def read_excel(*args, **kwargs) -> str:
     read_kwargs = kwargs_in_func(kwargs, pd.read_excel)
     df = pd.read_excel(*args, **read_kwargs)
 
     markdown_kwargs = kwargs_not_in_func(kwargs, pd.read_excel)
     return convert_to_md_table(df, markdown_kwargs)
 
 
-def read_yaml(*args, **kwargs):
+@ParseArgs
+def read_yaml(*args, **kwargs) -> str:
 
     json_kwargs = kwargs_in_func(kwargs, pd.json_normalize)
     with open(args[0], "r") as f:
@@ -54,14 +111,18 @@ def read_yaml(*args, **kwargs):
     markdown_kwargs = kwargs_not_in_func(kwargs, pd.json_normalize)
     return convert_to_md_table(df, markdown_kwargs)
 
-def read_feather(*args, **kwargs):
+
+@ParseArgs
+def read_feather(*args, **kwargs) -> str:
     read_kwargs = kwargs_in_func(kwargs, pd.read_feather)
     df = pd.read_feather(*args, **read_kwargs)
 
     markdown_kwargs = kwargs_not_in_func(kwargs, pd.read_feather)
     return convert_to_md_table(df, markdown_kwargs)
 
-def read_raw(*args, **kwargs):
+
+@ParseArgs
+def read_raw(*args, **kwargs) -> str:
     """Read a file as-is.
 
     Returns:

tests/fixtures/basic_setup/mkdocs_w_macros.yml

@@ -3,5 +3,5 @@ use_directory_urls: true
 
 plugins:
     - search
-    - table-reader
-    - macros
+    - macros
+    - table-reader

tests/fixtures/basic_setup/mkdocs_w_macros_wrong_order.yml

@@ -3,5 +3,5 @@ use_directory_urls: true
 
 plugins:
     - search
-    - macros
-    - table-reader
+    - table-reader
+    - macros

tests/fixtures/jinja/docs/basic_table.csv

@@ -0,0 +1,5 @@
+"a","b"
+40,73
+50,52
+531456,80
+"name","table1"

…sic_setup/assets/tables/basic_table2.csv …sts/fixtures/jinja/docs/basic_table2.csvtests/fixtures/basic_setup/assets/tables/basic_table2.csv renamed to tests/fixtures/jinja/docs/basic_table2.csv tests/fixtures/basic_setup/assets/tables/basic_table2.csv renamed to tests/fixtures/jinja/docs/basic_table2.csv

@@ -0,0 +1,15 @@
+# Test page
+
+This is a table that we load from the docs folder, because we set `data_path` to `docs`:
+
+## Dynamically insert a list of tables using jinja2
+
+
+{% set table_names = ["basic_table.csv","basic_table2.csv"] %}
+{% for table_name in table_names %}
+
+### table `{{ table_name }}`
+
+{{ read_csv(table_name) }}
+
+{% endfor %}

tests/fixtures/jinja/docs/index.md

@@ -0,0 +1,5 @@
+"a","b"
+40,73
+50,52
+531456,80
+"name","table1"

tests/fixtures/jinja/docs/subfolder/sub_table.csv

@@ -0,0 +1,5 @@
+# page
+
+Test that searching in page directory works
+
+{{ read_csv("sub_table.csv") }}

tests/fixtures/jinja/docs/subfolder/subpage.md

@@ -0,0 +1,7 @@
+site_name: test git_table_reader site
+use_directory_urls: true
+
+plugins:
+    - search
+    - macros
+    - table-reader