refactor tests and add more typing

26 files changed, 133 insertions, 178 deletions

diff --git a/README.md b/README.md
index f971860..4686564 100644
--- a/README.md
+++ b/README.md
@@ -12,19 +12,15 @@ Initially inspired by Roman Zolotarev's [`ssg5`](https://rgz.ee/bin/ssg5) and [`
 	- [x] Uses [`jinja`](https://jinja.palletsprojects.com/en/3.0.x/) for templating.
 	- [x] Preserves hand-made `*.html` files.
 	- [x] Tag functionality, useful for blog-style sites.
-	- [ ] Open Graph (and similar) support.
-		- Technically, this works if you add the correct metadata to the `*.md` files and use the variables available for Jinja.
 - [x] Build `sitemap.xml` file.
 	- [ ] Include manually added `*.html` files.
 - [x] Build `rss.xml` file.
-	- [ ] Join the `static_url` to all relative URLs found to comply with the [RSS 2.0 spec](https://validator.w3.org/feed/docs/rss2.html).
-		- This would be added to the parsed HTML text extracted from the MD files, so it would be available to the created `*.html` and `*.xml` files. Note that depending on the reader, it will append the URL specified in the RSS file or use the [`xml:base`](https://www.rssboard.org/news/151/relative-links) specified (for example, [newsboat](https://newsboat.org/) parses `xml:base`).
 	- [ ] Include manually added `*.html` files.
 - [x] YAML for configuration file, uses [`PyYAML`](https://pyyaml.org/).
-	- [ ] Handle multiple "documents".
-	- [ ] More complex directory structure to support multiple subdomains and different types of pages.
+	- [x] Handle multiple "documents". `PyYAML` supports this.
+	- [x] More complex directory structure to support multiple subdomains and different types of pages. This is supported by using mupltiple "documents" in the `yaml` config file.
+- [x] File checksum checking for modification of files.
 - [ ] Option/change to using an SQL database instead of the custom solution.
-- [x] Checksum checking because the timestamp of the file is not enough.
 - [ ] Use external markdown extensions.
 
 ### Markdown features
@@ -37,6 +33,7 @@ This program uses the base [`markdown` syntax](https://daringfireball.net/projec
 - SmartyPants.
 - Table of Contents. (With defaults as specified [here](https://python-markdown.github.io/extensions/toc/))
 - WikiLinks.
+- [pymdvar](https://github.com/luevano/pymdvar) (made by me).
 - [yafg - Yet Another Figure Generator](https://git.sr.ht/~ferruck/yafg)
 - [Markdown Checklist](https://github.com/FND/markdown-checklist)
 - [PyMdown Extensions](https://facelessuser.github.io/pymdown-extensions/)
@@ -196,19 +193,18 @@ These variables are exposed to use within the templates. The below list is displ
     - `toc` (`str`): table of contents as taken from `md.toc`.
     - `toc_tokens` (`list(dict)`): table of contents tokens as taken from `md.toc_tokens`.
 	- `cdatetime` (`datetime.datetime`): creation datetime object of the page.
-	- `cdate` (`method`): method thtat takes the name of the `fmt.FMT` and applies it to the `cdatetime` object.
+	- `cdate` (`method`): method that takes the name of the `fmt.FMT` and applies it to the `cdatetime` object.
 	- `cdate_rss` (`str`): formatted `cdatetime` as required by rss.
 	- `cdate_sitemap` (`str`): formatted `cdatetime` as required by sitemap.
 	- `mdatetime` (`datetime.datetime`): modification datetime object of the page. Defaults to `None`.
-	- `mdate` (`method`): method thtat takes the name of the `fmt.FMT` and applies it to the `mdatetime` object.
+	- `mdate` (`method`): method that takes the name of the `fmt.FMT` and applies it to the `mdatetime` object.
 	- `mdate_rss` (`str`): formatted `mdatetime` as required by rss.
 	- `mdate_sitemap` (`str`): formatted `mdatetime` as required by sitemap.
 	- `tags` (`list(tuple(str))`): list of tuple of tags of the page, containing the name and the url of the tag, in that order. Defaults to empty list.
 	- `url` (`str`): url of the page, this already includes the `url/main` from config file.
 	- `image_url` (`str`): image url of the page, this already includes the `url/static`. Defaults to the `url/default_image` config option.
 	- `next/previous` (`Page`): reference to the next or previous page object (containing all these attributes). Defaults to `None`.
-	- `og` (`dict(str, str)`): dict for object graph metadata.
-	- `meta` (`dict(str, list(str))`): meta dict as obtained from python-markdown, in case you use a meta tag not yet supported, it will be available there.
+	- `meta` (`dict(str, list(str))`): meta dict as obtained from python-markdown, in case you use a meta tag not directly supported, it will be available there.
 - `tag` (`tuple(str)`) (`tag.html`): tuple of name and url of the current tag.
 - `tag_pages` (`list(Page)`) (`tag.html`): similar to `all_pages` but contains all the pages for the current tag.
 - `all_tags` (`list(tuple(str))`) (all): similar to `page.tags` but contains all the tags.
diff --git a/requirements.txt b/requirements.txt
index 6078047..807946b 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -6,3 +6,4 @@ yafg>=0.3
 pymdown-extensions>=9.9.2
 pymdvar>=1.0.3
 PyYAML>=6.0
+validators>=0.20.0
diff --git a/setup.cfg b/setup.cfg
index 1fb6efd..bf05793 100644
--- a/setup.cfg
+++ b/setup.cfg
@@ -63,6 +63,7 @@ per-file-ignores =
 	arg_parser.py: E501
 	custom_logger.py: E501
 	test_database_entry.py: E501
+	test_configuration.py: E501
 
 [pbr]
 skip_authors = True
diff --git a/src/pyssg/configuration.py b/src/pyssg/configuration.py
index e2dc26b..c7b4248 100644
--- a/src/pyssg/configuration.py
+++ b/src/pyssg/configuration.py
@@ -12,6 +12,7 @@ DEFAULT_CONFIG_PATH: str = '$XDG_CONFIG_HOME/pyssg/config.yaml'
 VERSION: str = version('pyssg')
 
 
+# TODO: add checking for extensions config (such as pymdvar)
 def __check_well_formed_config(config: dict[str, Any],
                                config_base: list[dict[str, Any]],
                                prefix_key: str = '') -> None:
@@ -80,6 +81,7 @@ def get_static_config(sc_package: str = 'static_config.yaml',
     log.debug('reading and setting static config')
     config: dict[str, Any] = get_parsed_yaml(sc_package, plt_resource)[0]
 
+    # TODO: move this to utils and update the tests
     def __time(fmt: str) -> str:
         return datetime.now(tz=timezone.utc).strftime(config['fmt'][fmt])
 
diff --git a/src/pyssg/md_parser.py b/src/pyssg/md_parser.py
index 3ef297d..3b62985 100644
--- a/src/pyssg/md_parser.py
+++ b/src/pyssg/md_parser.py
@@ -4,6 +4,7 @@ from logging import Logger, getLogger
 
 from markdown import Markdown
 from yafg import YafgExtension
+from pymdvar import VariableExtension
 from markdown_checklist.extension import ChecklistExtension
 
 from .database import Database
@@ -12,13 +13,17 @@ from .page import Page
 log: Logger = getLogger(__name__)
 
 
-def _get_md_obj() -> Markdown:
+# TODO: add configuration testing for extensions config (pymdvar for ex)
+def get_md_obj(variables: dict[str, str] = dict(),
+               enable_env: bool = False) -> Markdown:
     exts: list = ['extra',
                   'meta',
                   'sane_lists',
                   'smarty',
                   'toc',
                   'wikilinks',
+                  VariableExtension(variables=variables,
+                                    enable_env=enable_env),
                   # stripTitle generates an error when True,
                   # if there is no title attr
                   YafgExtension(stripTitle=False,
@@ -50,7 +55,7 @@ class MDParser:
         self.config: dict = config
         self.dir_config: dict = dir_config
         self.db: Database = db
-        self.md: Markdown = _get_md_obj()
+        self.md: Markdown = get_md_obj()
 
         self.all_files: list[Page] = []
         self.all_tags: list[tuple[str, str]] = []
@@ -65,7 +70,8 @@ class MDParser:
 
             log.debug('parsing md into html')
             content: str = self.md.reset().convert(open(src_file).read())
-            # ignoring md.Meta type as it is not yet defined (because it is from an extension)
+            # ignoring md.Meta type as it is not yet defined
+            #   (because it is from an extension)
             page: Page = Page(f,
                               self.db.e[f].ctimestamp,
                               self.db.e[f].mtimestamp,
diff --git a/src/pyssg/page.py b/src/pyssg/page.py
index 6b8916d..4902bea 100644
--- a/src/pyssg/page.py
+++ b/src/pyssg/page.py
@@ -1,22 +1,21 @@
-import sys
 from datetime import datetime, timezone
 from logging import Logger, getLogger
+from typing import Any
 
 log: Logger = getLogger(__name__)
 
 
 class Page:
-    def __init__(self,
-                 name: str,
-                 ctime: float,
-                 mtime: float,
-                 html: str,
-                 toc: str,
-                 toc_tokens: list[str],
-                 meta: dict,
-                 config: dict,
-                 dir_config: dict) -> None:
-        log.debug('initializing the page object with name "%s"', name)
+    def __init__(self, name: str,
+                       ctime: float,
+                       mtime: float,
+                       html: str,
+                       toc: str,
+                       toc_tokens: list[str],
+                       meta: dict[str, Any],
+                       config: dict[str, Any],
+                       dir_config: dict[str, Any]) -> None:
+        log.debug('initializing a page object with name "%s"', name)
         # initial data
         self.name: str = name
         self.ctimestamp: float = ctime
@@ -24,9 +23,9 @@ class Page:
         self.content: str = html
         self.toc: str = toc
         self.toc_tokens: list[str] = toc_tokens
-        self.meta: dict = meta
-        self.config: dict = config
-        self.dir_config: dict = dir_config
+        self.meta: dict[str, Any] = meta
+        self.config: dict[str, Any] = config
+        self.dir_config: dict[str, Any] = dir_config
 
         # data from self.meta
         self.title: str
@@ -48,25 +47,25 @@ class Page:
         self.next: Page | None = None
         self.previous: Page | None = None
 
-        # also from self.meta, but for og metadata
-        self.og: dict[str, str] = dict()
-
     def __lt__(self, other):
         return self.ctimestamp < other.ctimestamp
 
-    def __get_meta(self, var: str, or_else: str | list[str]) -> str | list[str]:
+    def __get_meta(self, var: str,
+                   or_else: str | list[str] = '') -> str | list[str] | Any:
         if var in self.meta:
             log.debug('getting metadata "%s"', var)
             return self.meta[var]
         else:
-            log.debug('getting metadata "%s" failed, using optional value "%s"', var, or_else)
+            log.debug('getting metadata "%s" failed, using optional value "%s"',
+                      var, or_else)
             return or_else
 
     def cdate(self, format: str) -> str:
         if format in self.config['fmt']:
             return self.cdatetime.strftime(self.config['fmt'][format])
         else:
-            log.warning('format "%s" not found in config["fmt"], returning empty string', format)
+            log.warning('format "%s" not found in config, returning '
+                        'empty string', format)
             return ''
 
     def mdate(self, format: str) -> str:
@@ -74,28 +73,32 @@ class Page:
             log.warning('no mdatetime found, can\'t return a formatted string')
             return ''
         if format in self.config['fmt']:
-            return self.mdatetime.strftime(self.config['fmt'][format])  # type: ignore
+            return self.mdatetime.strftime(self.config['fmt'][format])
         else:
-            log.warning('format "%s" not found in config["fmt"], returning empty string', format)
+            log.warning('format "%s" not found in config, returning '
+                        'empty string', format)
             return ''
 
+    def from_timestamp(self, timestamp: float) -> datetime:
+        return datetime.fromtimestamp(timestamp, tz=timezone.utc)
+
     # parses meta from self.meta, for og, it prioritizes,
     #   the actual og meta
     def parse_metadata(self):
         log.debug('parsing metadata for file "%s"', self.name)
-        self.title = self.__get_meta('title', [''])[0]
+        self.title = str(self.__get_meta('title'))
         self.author = list(self.__get_meta('author', ['']))
-        self.summary = self.__get_meta('summary', [''])[0]
-        self.lang = self.__get_meta('lang', ['en'])[0]
+        self.summary = str(self.__get_meta('summary'))
+        self.lang = str(self.__get_meta('lang', 'en'))
 
         log.debug('parsing timestamp')
-        self.cdatetime = datetime.fromtimestamp(self.ctimestamp, tz=timezone.utc)
+        self.cdatetime = self.from_timestamp(self.ctimestamp)
         self.cdate_rss = self.cdate('rss_date')
         self.cdate_sitemap = self.cdate('sitemap_date')
 
         if self.mtimestamp != 0.0:
             log.debug('parsing modified timestamp')
-            self.mdatetime = datetime.fromtimestamp(self.mtimestamp, tz=timezone.utc)
+            self.mdatetime = self.from_timestamp(self.mtimestamp)
             self.mdate_rss = self.mdate('rss_date')
             self.mdate_sitemap = self.mdate('sitemap_date')
         else:
@@ -108,30 +111,35 @@ class Page:
                 tags_only.sort()
 
                 for t in tags_only:
-                    # need to specify dir_config['url'] as it is a hardcoded tag url
-                    self.tags.append((t, f'{self.dir_config["url"]}/tag/@{t}.html'))
+                    # need to specify dir_config['url'] as it is
+                    #   a hardcoded tag url
+                    tag_url: str = f'{self.dir_config["url"]}/tag/@{t}.html'
+                    self.tags.append((t, tag_url))
             else:
                 log.debug('no tags to parse')
 
-        log.debug('parsing url')
-        # no need to specify dir_config['url'] as self.name already contains the relative url
-        self.url = f'{self.config["url"]["main"]}/{self.name.replace(".md", ".html")}'
+        log.debug('parsing page url')
+        # no need to specify dir_config['url'] as self.name already
+        #   contains the relative url
+        name_html: str = self.name.replace(".md", ".html")
+        self.url = f'{self.config["url"]["main"]}/{name_html}'
         log.debug('final url "%s"', self.url)
 
         log.debug('parsing image url')
         default_image_url: str = ''
         if 'default_image' in self.config['url']:
-            log.debug('"default_image" url found, will use if no "image_url" is found')
+            log.debug('"default_image" url found, will use if no "image_url" '
+                      'is found')
             default_image_url = self.config['url']['default_image']
 
         image_url: str
-        image_url = self.__get_meta('image_url', [default_image_url])[0]
+        image_url = str(self.__get_meta('image_url', default_image_url))
 
         if image_url != '':
             if 'static' in self.config['url']:
                 self.image_url = f'{self.config["url"]["static"]}/{image_url}'
             else:
-                log.debug('no static url set, using main url, this could cause problems')
+                log.debug('no static url set, using main url')
                 self.image_url = f'{self.config["url"]["main"]}/{image_url}'
             log.debug('final image url "%s"', self.image_url)
         else:
@@ -139,24 +147,3 @@ class Page:
             log.debug('no image url set for the page, could be because no'
                       ' "image_url" was found in the metadata and/or no '
                       ' "default_image" set in the config file')
-
-        # if contains open graph elements
-        # TODO: better handle this part
-        # og_e = object graph entry
-        og_elements: list[str] = list(self.__get_meta('og', []))
-        if og_elements:
-            log.debug('parsing og metadata')
-            for og_e in og_elements:
-                kv: list[str] = og_e.split(',', 1)
-                if len(kv) != 2:
-                    log.error('invalid og syntax for "%s", needs to be "k, v"', og_e)
-                    sys.exit(1)
-
-                k: str = kv[0].strip()
-                v: str = kv[1].strip()
-
-                log.debug('og element: ("%s", "%s")', k, v)
-                self.og[k] = v
-
-        else:
-            log.debug('no tags to parse')
diff --git a/src/pyssg/plt/default.yaml b/src/pyssg/plt/default.yaml
index 0b722a6..ca2f7ad 100644
--- a/src/pyssg/plt/default.yaml
+++ b/src/pyssg/plt/default.yaml
@@ -10,12 +10,8 @@ path:
   db: !join [*root, ".files"]
 url:
   main: "https://example.com"
-  static: "https://static.example.com"
-  default_image: "images/default.png"
 fmt:
   date: "%a, %b %d, %Y @ %H:%M %Z"
-  list_date: "%b %d"
-  list_sep_date: "%B %Y"
 dirs:
   /:
     cfg:
@@ -24,5 +20,4 @@ dirs:
       index: False
       rss: False
       sitemap: False
-      exclude_dirs: []
 ...
\ No newline at end of file
diff --git a/tests/conftest.py b/tests/conftest.py
index d3f28d7..aaf0b3a 100644
--- a/tests/conftest.py
+++ b/tests/conftest.py
@@ -31,18 +31,18 @@ def sitemap_date_fmt():
 
 
 @pytest.fixture(scope='session')
-def test_dir() -> str:
-    return str(os.path.dirname(os.path.abspath(__file__)))
+def sample_files_path() -> str:
+    return f'{str(os.path.dirname(os.path.abspath(__file__)))}/sample_files'
 
 
 @pytest.fixture(scope='session')
-def test_resource() -> str:
-    return 'tests.io_files'
+def config_resource() -> str:
+    return 'tests.sample_files.config'
 
 
 @pytest.fixture(scope='session')
-def simple_yaml() -> str:
-    return 'simple.yaml'
+def default_yaml() -> str:
+    return 'default.yaml'
 
 
 @pytest.fixture(scope='session')
@@ -76,7 +76,7 @@ def get_fmt_time() -> Callable[..., str]:
 
 
 @pytest.fixture
-def simple_dict() -> dict[str, Any]:
+def default_config_dict() -> dict[str, Any]:
     return {'define': '$PYSSG_HOME/pyssg/site_example/',
             'title': 'Example site',
             'path': {
@@ -85,13 +85,9 @@ def simple_dict() -> dict[str, Any]:
                 'plt': '/tmp/pyssg/pyssg/site_example/plt',
                 'db': '/tmp/pyssg/pyssg/site_example/.files'},
             'url': {
-                'main': 'https://example.com',
-                'static': 'https://static.example.com',
-                'default_image': 'images/default.png'},
+                'main': 'https://example.com'},
             'fmt': {
-                'date': '%a, %b %d, %Y @ %H:%M %Z',
-                'list_date': '%b %d',
-                'list_sep_date': '%B %Y'},
+                'date': '%a, %b %d, %Y @ %H:%M %Z'},
             'dirs': {
                 '/': {
                     'cfg': {
@@ -99,13 +95,12 @@ def simple_dict() -> dict[str, Any]:
                         'tags': False,
                         'index': False,
                         'rss': False,
-                        'sitemap': False,
-                        'exclude_dirs': []}}}}
+                        'sitemap': False}}}}
 
 
 @pytest.fixture(scope='function')
 def tmp_dir_structure(tmp_path: Path) -> Path:
-    root: Path = tmp_path/'dir_str'
+    root: Path = tmp_path/'dir_structure'
     # order matters
     dirs: list[Path] = [root,
                         root/'first',
@@ -165,12 +160,12 @@ def tmp_db_wrong_col_num(tmp_path: Path) -> Path:
 
 @pytest.fixture(scope='function')
 def tmp_src_dir(tmp_path: Path,
-                test_dir: str) -> Path:
+                sample_files_path: str) -> Path:
     src: Path = tmp_path/'src'
     src_a: Path = src/'a'
     src.mkdir()
     src_a.mkdir()
-    src_test: str = f'{test_dir}/io_files/md'
+    src_test: str = f'{sample_files_path}/md'
 
     files: list[str] = ['first.md', 'new.md', 'a/second.md']
     for f in files:
diff --git a/tests/io_files/__init__.py b/tests/sample_files/__init__.py
index e69de29..e69de29 100644
--- a/tests/io_files/__init__.py
+++ b/tests/sample_files/__init__.py
diff --git a/tests/sample_files/checksum.txt b/tests/sample_files/checksum.txt
new file mode 100644
index 0000000..025b879
--- /dev/null
+++ b/tests/sample_files/checksum.txt
@@ -0,0 +1 @@
+The content of this file is irrelevant as it is only to test the checksum function.
\ No newline at end of file
diff --git a/tests/io_files/md/__init__.py b/tests/sample_files/config/__init__.py
index e69de29..e69de29 100644
--- a/tests/io_files/md/__init__.py
+++ b/tests/sample_files/config/__init__.py
diff --git a/tests/io_files/simple.yaml b/tests/sample_files/config/default.yaml
index df3888b..08121a6 100644
--- a/tests/io_files/simple.yaml
+++ b/tests/sample_files/config/default.yaml
@@ -10,12 +10,8 @@ path:
   db: !join [*root, ".files"]
 url:
   main: "https://example.com"
-  static: "https://static.example.com"
-  default_image: "images/default.png"
 fmt:
   date: "%a, %b %d, %Y @ %H:%M %Z"
-  list_date: "%b %d"
-  list_sep_date: "%B %Y"
 dirs:
   /:
     cfg:
@@ -24,5 +20,4 @@ dirs:
       index: False
       rss: False
       sitemap: False
-      exclude_dirs: []
 ...
\ No newline at end of file
diff --git a/tests/io_files/simple_missing_dirs.yaml b/tests/sample_files/config/default_missing_dirs.yaml
index aa15fb5..03ee35a 100644
--- a/tests/io_files/simple_missing_dirs.yaml
+++ b/tests/sample_files/config/default_missing_dirs.yaml
@@ -10,12 +10,8 @@ path:
   db: !join [*root, ".files"]
 url:
   main: "https://example.com"
-  static: "https://static.example.com"
-  default_image: "images/default.png"
 fmt:
   date: "%a, %b %d, %Y @ %H:%M %Z"
-  list_date: "%b %d"
-  list_sep_date: "%B %Y"
 dirs:
-# test missing dirs
+# test missing dirs (doesn't have any)
 ...
\ No newline at end of file
diff --git a/tests/io_files/simple_missing_key.yaml b/tests/sample_files/config/default_missing_mandatory_key.yaml
index ac81563..b5554f7 100644
--- a/tests/io_files/simple_missing_key.yaml
+++ b/tests/sample_files/config/default_missing_mandatory_key.yaml
@@ -11,12 +11,8 @@ path:
   db: !join [*root, ".files"]
 url:
   main: "https://example.com"
-  static: "https://static.example.com"
-  default_image: "images/default.png"
 fmt:
   date: "%a, %b %d, %Y @ %H:%M %Z"
-  list_date: "%b %d"
-  list_sep_date: "%B %Y"
 dirs:
   /:
     cfg:
@@ -25,5 +21,4 @@ dirs:
       index: False
       rss: False
       sitemap: False
-      exclude_dirs: []
 ...
\ No newline at end of file
diff --git a/tests/io_files/simple_missing_root_dir.yaml b/tests/sample_files/config/default_missing_root_dir.yaml
index 07fa824..896e141 100644
--- a/tests/io_files/simple_missing_root_dir.yaml
+++ b/tests/sample_files/config/default_missing_root_dir.yaml
@@ -10,13 +10,9 @@ path:
   db: !join [*root, ".files"]
 url:
   main: "https://example.com"
-  static: "https://static.example.com"
-  default_image: "images/default.png"
 fmt:
   date: "%a, %b %d, %Y @ %H:%M %Z"
-  list_date: "%b %d"
-  list_sep_date: "%B %Y"
 dirs:
-# test missing /:
+# test missing "/" dir in specific
   something:
 ...
\ No newline at end of file
diff --git a/tests/io_files/multiple.yaml b/tests/sample_files/config/multiple_default.yaml
index 8d99c40..54954b1 100644
--- a/tests/io_files/multiple.yaml
+++ b/tests/sample_files/config/multiple_default.yaml
@@ -10,12 +10,8 @@ path:
   db: !join [*root, ".files"]
 url:
   main: "https://example.com"
-  static: "https://static.example.com"
-  default_image: "images/default.png"
 fmt:
   date: "%a, %b %d, %Y @ %H:%M %Z"
-  list_date: "%b %d"
-  list_sep_date: "%B %Y"
 dirs:
   /:
     cfg:
@@ -24,7 +20,6 @@ dirs:
       index: False
       rss: False
       sitemap: False
-      exclude_dirs: []
 ...
 ---
 define: &root "$PYSSG_HOME/pyssg/site_example/"
@@ -37,12 +32,8 @@ path:
   db: !join [*root, ".files"]
 url:
   main: "https://example.com"
-  static: "https://static.example.com"
-  default_image: "images/default.png"
 fmt:
   date: "%a, %b %d, %Y @ %H:%M %Z"
-  list_date: "%b %d"
-  list_sep_date: "%B %Y"
 dirs:
   /:
     cfg:
@@ -51,5 +42,4 @@ dirs:
       index: False
       rss: False
       sitemap: False
-      exclude_dirs: []
 ...
\ No newline at end of file
diff --git a/tests/io_files/multiple_one_doc_error.yaml b/tests/sample_files/config/multiple_default_one_doc_error.yaml
index 86f6546..44d9beb 100644
--- a/tests/io_files/multiple_one_doc_error.yaml
+++ b/tests/sample_files/config/multiple_default_one_doc_error.yaml
@@ -10,12 +10,8 @@ path:
   db: !join [*root, ".files"]
 url:
   main: "https://example.com"
-  static: "https://static.example.com"
-  default_image: "images/default.png"
 fmt:
   date: "%a, %b %d, %Y @ %H:%M %Z"
-  list_date: "%b %d"
-  list_sep_date: "%B %Y"
 dirs:
   /:
     cfg:
@@ -24,7 +20,6 @@ dirs:
       index: False
       rss: False
       sitemap: False
-      exclude_dirs: []
 ...
 ---
 define: &root "$PYSSG_HOME/pyssg/site_example/"
@@ -37,12 +32,8 @@ path:
   db: !join [*root, ".files"]
 url:
   main: "https://example.com"
-  static: "https://static.example.com"
-  default_image: "images/default.png"
 fmt:
   date: "%a, %b %d, %Y @ %H:%M %Z"
-  list_date: "%b %d"
-  list_sep_date: "%B %Y"
 dirs:
   # just removing all paths as it will cause an error
 ...
\ No newline at end of file
diff --git a/tests/io_files/md/a/__init__.py b/tests/sample_files/md/__init__.py
index e69de29..e69de29 100644
--- a/tests/io_files/md/a/__init__.py
+++ b/tests/sample_files/md/__init__.py
diff --git a/tests/sample_files/md/a/__init__.py b/tests/sample_files/md/a/__init__.py
new file mode 100644
index 0000000..e69de29
--- /dev/null
+++ b/tests/sample_files/md/a/__init__.py
diff --git a/tests/io_files/md/a/second.md b/tests/sample_files/md/a/second.md
index cb4e333..cb4e333 100644
--- a/tests/io_files/md/a/second.md
+++ b/tests/sample_files/md/a/second.md
diff --git a/tests/io_files/md/first.md b/tests/sample_files/md/first.md
index 567ea3e..567ea3e 100644
--- a/tests/io_files/md/first.md
+++ b/tests/sample_files/md/first.md
diff --git a/tests/io_files/md/new.md b/tests/sample_files/md/new.md
index ce684a7..ce684a7 100644
--- a/tests/io_files/md/new.md
+++ b/tests/sample_files/md/new.md
diff --git a/tests/test_configuration.py b/tests/test_configuration.py
index 68f7808..36761eb 100644
--- a/tests/test_configuration.py
+++ b/tests/test_configuration.py
@@ -7,10 +7,10 @@ from pyssg.configuration import get_static_config, get_parsed_config
 
 # this test is a bit sketchy, as the way the datetimes are calculated could vary
 #   by milliseconds or even have a difference in seconds
-def test_static_default(rss_date_fmt: str,
-                        sitemap_date_fmt: str,
-                        get_fmt_time: Callable[..., str],
-                        version: str) -> None:
+def test_static_config(rss_date_fmt: str,
+                       sitemap_date_fmt: str,
+                       get_fmt_time: Callable[..., str],
+                       version: str) -> None:
     rss_run_date: str = get_fmt_time(rss_date_fmt)
     sitemap_run_date: str = get_fmt_time(sitemap_date_fmt)
     sc_dict: dict[str, Any] = {'fmt': {'rss_date': rss_date_fmt,
@@ -22,21 +22,21 @@ def test_static_default(rss_date_fmt: str,
     assert static_config == sc_dict
 
 
-def test_simple(test_dir: str,
-                simple_yaml: str,
-                simple_dict: dict[str, Any]) -> None:
-    yaml_path: str = f'{test_dir}/io_files/{simple_yaml}'
+def test_default_config(sample_files_path: str,
+                        default_yaml: str,
+                        default_config_dict: dict[str, Any]) -> None:
+    yaml_path: str = f'{sample_files_path}/config/{default_yaml}'
     yaml: list[dict[str, Any]] = get_parsed_config(yaml_path)
     assert len(yaml) == 1
-    assert yaml[0] == simple_dict
+    assert yaml[0] == default_config_dict
 
 
-def test_simple_mising_key(test_dir: str,
-                           caplog: LogCaptureFixture) -> None:
+def test_default_config_mising_mandatory_key(sample_files_path: str,
+                                             caplog: LogCaptureFixture) -> None:
     err: tuple[str, int, str] = ('pyssg.configuration',
                                  ERROR,
                                  'config doesn\'t have "title"')
-    yaml_path: str = f'{test_dir}/io_files/simple_missing_key.yaml'
+    yaml_path: str = f'{sample_files_path}/config/default_missing_mandatory_key.yaml'
     with pytest.raises(SystemExit) as system_exit:
         get_parsed_config(yaml_path)
     assert system_exit.type == SystemExit
@@ -44,12 +44,12 @@ def test_simple_mising_key(test_dir: str,
     assert caplog.record_tuples[-1] == err
 
 
-def test_simple_mising_dirs(test_dir: str,
-                            caplog: LogCaptureFixture) -> None:
+def test_default_config_mising_dirs(sample_files_path: str,
+                                    caplog: LogCaptureFixture) -> None:
     err: tuple[str, int, str] = ('pyssg.configuration',
                                  ERROR,
                                  'config doesn\'t have any dirs (dirs.*)')
-    yaml_path: str = f'{test_dir}/io_files/simple_missing_dirs.yaml'
+    yaml_path: str = f'{sample_files_path}/config/default_missing_dirs.yaml'
     with pytest.raises(SystemExit) as system_exit:
         get_parsed_config(yaml_path)
     assert system_exit.type == SystemExit
@@ -57,12 +57,12 @@ def test_simple_mising_dirs(test_dir: str,
     assert caplog.record_tuples[-1] == err
 
 
-def test_simple_root_dir(test_dir: str,
-                         caplog: LogCaptureFixture) -> None:
+def test_default_config_root_dir(sample_files_path: str,
+                                 caplog: LogCaptureFixture) -> None:
     err: tuple[str, int, str] = ('pyssg.configuration',
                                  ERROR,
                                  'config doesn\'t have "dirs./"')
-    yaml_path: str = f'{test_dir}/io_files/simple_missing_root_dir.yaml'
+    yaml_path: str = f'{sample_files_path}/config/default_missing_root_dir.yaml'
     with pytest.raises(SystemExit) as system_exit:
         get_parsed_config(yaml_path)
     assert system_exit.type == SystemExit
@@ -71,24 +71,24 @@ def test_simple_root_dir(test_dir: str,
 
 
 # this really just tests that both documents in the yaml file are read,
-#   multiple.yaml is just simple.yaml with the same document twice,
-#   shouldn't be an issue as the yaml package handles this
-def test_multiple(test_dir: str, simple_dict: dict[str, Any]) -> None:
-    yaml_path: str = f'{test_dir}/io_files/multiple.yaml'
+#   both documents are the same (the default.yaml)
+def test_multiple_default_config(sample_files_path: str,
+                                 default_config_dict: dict[str, Any]) -> None:
+    yaml_path: str = f'{sample_files_path}/config/multiple_default.yaml'
     yaml: list[dict[str, Any]] = get_parsed_config(yaml_path)
     assert len(yaml) == 2
-    assert yaml[0] == simple_dict
-    assert yaml[1] == simple_dict
+    assert yaml[0] == default_config_dict
+    assert yaml[1] == default_config_dict
 
 
 # also, this just tests that the checks for a well formed config file are
 #   processed for all documents
-def test_multiple_one_doc_error(test_dir: str,
-                                caplog: LogCaptureFixture) -> None:
+def test_multiple_default_config_one_doc_error(sample_files_path: str,
+                                               caplog: LogCaptureFixture) -> None:
     err: tuple[str, int, str] = ('pyssg.configuration',
                                  ERROR,
                                  'config doesn\'t have any dirs (dirs.*)')
-    yaml_path: str = f'{test_dir}/io_files/multiple_one_doc_error.yaml'
+    yaml_path: str = f'{sample_files_path}/config/multiple_default_one_doc_error.yaml'
     with pytest.raises(SystemExit) as system_exit:
         get_parsed_config(yaml_path)
     assert system_exit.type == SystemExit
diff --git a/tests/test_database.py b/tests/test_database.py
index 7ca597f..c5957e4 100644
--- a/tests/test_database.py
+++ b/tests/test_database.py
@@ -6,8 +6,9 @@ from pyssg.database import Database
 from pyssg.database_entry import DatabaseEntry
 
 
-def test_read_database_no_db(test_dir: str, caplog: LogCaptureFixture) -> None:
-    path: str = f'{test_dir}/non_existent_db.psv'
+def test_read_database_no_db(sample_files_path: str,
+                             caplog: LogCaptureFixture) -> None:
+    path: str = f'{sample_files_path}/non_existent_db.psv'
     war: tuple[str, int, str] = ('pyssg.database',
                                  WARNING,
                                  f'"{path}" doesn\'t exist, will be created '
@@ -18,9 +19,9 @@ def test_read_database_no_db(test_dir: str, caplog: LogCaptureFixture) -> None:
     assert caplog.record_tuples[-1] == war
 
 
-def test_read_database_not_a_file(test_dir: str,
+def test_read_database_not_a_file(sample_files_path: str,
                                   caplog: LogCaptureFixture) -> None:
-    path: str = test_dir
+    path: str = sample_files_path
     err: tuple[str, int, str] = ('pyssg.database',
                                  ERROR,
                                  f'"{path}" is not a file')
diff --git a/tests/test_utils.py b/tests/test_utils.py
index 86242c2..75b79c2 100644
--- a/tests/test_utils.py
+++ b/tests/test_utils.py
@@ -33,13 +33,14 @@ def test_path_expansion_failure(path: str) -> None:
     assert system_exit.value.code == 1
 
 
-def test_checksum(test_dir: str, simple_yaml: str) -> None:
-    path: str = f'{test_dir}/io_files/{simple_yaml}'
-    simple_yaml_checksum: str = 'd4f0a3ed56fd530d3ea485dced25534c'
+def test_checksum(sample_files_path: str) -> None:
+    path: str = f'{sample_files_path}/checksum.txt'
+    simple_yaml_checksum: str = '437b5a0e20d32fc14944c1c00d066303'
     checksum: str = get_checksum(path)
     assert checksum == simple_yaml_checksum
 
 
+# TODO: actually check the existence of the files and not just the log
 def test_copy_file(tmp_path: Path, caplog: LogCaptureFixture) -> None:
     src: Path = tmp_path/'src'
     dst: Path = tmp_path/'dst'
@@ -55,7 +56,9 @@ def test_copy_file(tmp_path: Path, caplog: LogCaptureFixture) -> None:
     assert caplog.record_tuples[-1] == inf
 
 
-def test_copy_file_failure(tmp_path: Path, caplog: LogCaptureFixture) -> None:
+# TODO: actually check the existence of the files and not just the log
+def test_copy_file_already_exists(tmp_path: Path,
+                                  caplog: LogCaptureFixture) -> None:
     src: Path = tmp_path/'src'
     dst: Path = tmp_path/'dst'
     src.mkdir()
@@ -82,7 +85,9 @@ def test_create_dir(tmp_path: Path, caplog: LogCaptureFixture) -> None:
     assert caplog.record_tuples[-1] == inf
 
 
-def test_create_dir_failure(tmp_path: Path, caplog: LogCaptureFixture) -> None:
+# TODO: actually check the existence of the files and not just the log
+def test_create_dir_already_exists(tmp_path: Path,
+                                   caplog: LogCaptureFixture) -> None:
     path: Path = tmp_path/'new_dir'
     inf: tuple[str, int, str] = ('pyssg.utils',
                                  INFO,
@@ -106,7 +111,9 @@ def test_create_dirs(tmp_path: Path, caplog: LogCaptureFixture) -> None:
     assert caplog.record_tuples[-1] == inf
 
 
-def test_create_dirs_failure(tmp_path: Path, caplog: LogCaptureFixture) -> None:
+# TODO: actually check the existence of the files and not just the log
+def test_create_dirs_already_exists(tmp_path: Path,
+                                    caplog: LogCaptureFixture) -> None:
     path: Path = tmp_path/'new_dir'
     sub_path: Path = path/'sub_dir'
     inf: tuple[str, int, str] = ('pyssg.utils',
@@ -131,7 +138,7 @@ def test_dir_structure(tmp_dir_structure: Path,
                        exclude: list[str],
                        exp_dir_str: list[str]) -> None:
     dir_str: list[str] = get_dir_structure(str(tmp_dir_structure), exclude)
-    # order doesn't matter
+    # order doesn't matter, only for checking that both lists contain the same
     assert sorted(dir_str) == sorted(exp_dir_str)
 
 
@@ -158,5 +165,5 @@ def test_file_list(tmp_dir_structure: Path,
                    exclude_dirs: list[str],
                    exp_flist: list[str]) -> None:
     flist: list[str] = get_file_list(str(tmp_dir_structure), exts, exclude_dirs)
-    # order doesn't matter
+    # order doesn't matter, only for checking that both lists contain the same
     assert sorted(flist) == sorted(exp_flist)
diff --git a/tests/test_yaml_parser.py b/tests/test_yaml_parser.py
index 906c7e6..0d8df96 100644
--- a/tests/test_yaml_parser.py
+++ b/tests/test_yaml_parser.py
@@ -5,19 +5,19 @@ from pyssg.yaml_parser import get_parsed_yaml
 #   and test the join functionality
 
 
-def test_yaml_resource_read(simple_yaml: str, test_resource: str) -> None:
-    yaml: list[dict[str, Any]] = get_parsed_yaml(simple_yaml, test_resource)
+def test_yaml_resource_read(default_yaml: str, config_resource: str) -> None:
+    yaml: list[dict[str, Any]] = get_parsed_yaml(default_yaml, config_resource)
     assert len(yaml) == 1
 
 
-def test_yaml_path_read(test_dir: str) -> None:
-    yaml_path: str = f'{test_dir}/io_files/simple.yaml'
+def test_yaml_path_read(sample_files_path: str, default_yaml: str) -> None:
+    yaml_path: str = f'{sample_files_path}/config/{default_yaml}'
     yaml: list[dict[str, Any]] = get_parsed_yaml(yaml_path)
     assert len(yaml) == 1
 
 
-def test_yaml_join(simple_yaml: str, test_resource: str) -> None:
-    yaml: dict[str, Any] = get_parsed_yaml(simple_yaml, test_resource)[0]
+def test_yaml_join(default_yaml: str, config_resource: str) -> None:
+    yaml: dict[str, Any] = get_parsed_yaml(default_yaml, config_resource)[0]
     define_str: str = '$PYSSG_HOME/pyssg/site_example/'
     assert yaml['define'] == define_str
     assert yaml['path']['src'] == f'{define_str}src'