Develop

.gitignore

@@ -12,10 +12,10 @@ rslv/data
 *.so
 
 # Distribution / packaging
+dist/
 .Python
 build/
 develop-eggs/
-dist/
 downloads/
 eggs/
 .eggs/

.pre-commit-config.yaml

@@ -0,0 +1,15 @@
+repos:
+-   repo: https://github.com/pre-commit/pre-commit-hooks
+    rev: v2.3.0
+    hooks:
+    -   id: check-yaml
+    -   id: end-of-file-fixer
+    -   id: trailing-whitespace
+- repo: https://github.com/astral-sh/uv-pre-commit
+  # uv version.
+  rev: 0.8.14
+  hooks:
+    # Update the uv lockfile
+    - id: uv-lock
+    - id: uv-export
+      args: ["--no-hashes", "--no-dev", "--no-group", "docs", "--output-file=requirements.txt"]

README.md

@@ -2,9 +2,9 @@
 
 FastAPI implementation of an identifier resolver.
 
-The `rslv`  implementation provides a generic identifier resolution service that may be adapted 
-to support different schemes. The base requirement is that identifiers are of the 
-form `scheme:content` where `scheme` is a scheme name (e.g. "doi" or "ark") and `content` is the 
+The `rslv`  implementation provides a generic identifier resolution service that may be adapted
+to support different schemes. The base requirement is that identifiers are of the
+form `scheme:content` where `scheme` is a scheme name (e.g. "doi" or "ark") and `content` is the
 value of the identifier (e.g. `10.12345/foo` or `99999/fd99`).
 
 ## Operation
@@ -38,4 +38,4 @@ With `uvicorn` installed, a development instance of `rslv` can be started from t
 python rslv/app.py
 ```
 
-The service may be accessed at http://localhost:8000/
+The service may be accessed at http://localhost:8000/

_docsrc/build.sh

@@ -57,4 +57,4 @@ codebraid \
   --wrap=none \
   --output "$dest" \
   --overwrite \
-  "$src"
+  "$src"

_docsrc/configuration.md

@@ -7,7 +7,7 @@ title: Configuring rslv
 ## Service Setup
 
 - `config.py` options
-- `logging.conf` 
+- `logging.conf`
 
 ## Resolver Configuration
 
@@ -55,5 +55,3 @@ scheme:prefix/value
 Target Template
 
 The target is specified in the definition as a template with placeholders that are filled by components of the parsed identifier.
-
-

_docsrc/doc_parts/__init__.py

@@ -108,4 +108,3 @@ def defn_match_table(pids=EXAMPLE_PIDS):
             }
             results.append(result)
         print(markdown_table(results).set_params(row_sep='markdown', quote=False).get_markdown())
-

_docsrc/index.md

@@ -4,7 +4,7 @@ title: "`rslv` Generic Resolver Service"
 
 `rslv` implements a resolver service. That is, given an identifier string, the service returns information about the identifier or redirects to the known location of the identified resource.
 
-`rslv` is written in Python and requires python version 3.9 or later. `rslv` may be run as a command line application or more typically, as a web service. 
+`rslv` is written in Python and requires python version 3.9 or later. `rslv` may be run as a command line application or more typically, as a web service.
 
 ## Installation
 
@@ -36,7 +36,7 @@ For development purposes, `rslv` may be run from the command line to provide a t
 python rslv/app.py
 ```
 
-A production deployment should use an ASGI server such as [Uvicorn](https://www.uvicorn.org/) or [Nginx Unit](https://unit.nginx.org/). Uivcorn will generally be deployed behind another web server such as Apache or Nginx whereas `Unit` may be deployed as the web server. 
+A production deployment should use an ASGI server such as [Uvicorn](https://www.uvicorn.org/) or [Nginx Unit](https://unit.nginx.org/). Uivcorn will generally be deployed behind another web server such as Apache or Nginx whereas `Unit` may be deployed as the web server.
 
 
 
@@ -45,4 +45,4 @@ Alternatively, `rslv` may be deployed to a cloud provider such as `Vercel`
 
 
 
-## 
+##

_docsrc/matching.md

@@ -22,7 +22,7 @@ It does this by splitting the input identifier string into various components an
 
 <figcaption>
 
-**Figure 1.** Overview of process for handling a user supplied identifier string. The string is split into components as a `parsed_pid` instance. That instance is matched against the available definitions. A match provides a `pid_definition` instance which is used with the `If a match is found then the response is a redirect to the registered target or the matched definition metadata. 
+**Figure 1.** Overview of process for handling a user supplied identifier string. The string is split into components as a `parsed_pid` instance. That instance is matched against the available definitions. A match provides a `pid_definition` instance which is used with the `If a match is found then the response is a redirect to the registered target or the matched definition metadata.
 
 </figcaption>
 
@@ -38,7 +38,7 @@ The provided identifier string is split into several components (Figure 2) by ap
 3. Left trim whitespace or any instances of the characters `:`, `/` from the second portion. This portion is the `content`.
 4. Split `content` at the first occurrence of the forward slash character ("/").
 5. The first portion is the `prefix`
-6. Left trim whitespace pr any instance of the characters `:`, `/` from the second portion. This portion is the `value`. 
+6. Left trim whitespace pr any instance of the characters `:`, `/` from the second portion. This portion is the `value`.
 
 <figure>
 
@@ -50,7 +50,7 @@ identifier = ark:12345/some_value/with?extra=foo
               |    |             |
            scheme  |           value
                  prefix
-       
+
 scheme = "ark"
 content = "12345/some_value/with?extra=foo"
 prefix = "12345"
@@ -59,7 +59,7 @@ value = "some_value/with?extra=foo"
 
 <figcaption>
 
-**Figure 2.** Components of a `parsed_pid`. After parsing, extracted components of the identifier are available for locating a matching definition and formatting the response. 
+**Figure 2.** Components of a `parsed_pid`. After parsing, extracted components of the identifier are available for locating a matching definition and formatting the response.
 
 </figcaption>
 </figure>
@@ -105,7 +105,7 @@ examples = [
   "ark:99999/foozle",
   "ark:example/foozle",
   "ark:99999/fk4qwerty",
-  "ark:99999/fkqwerty",  
+  "ark:99999/fkqwerty",
 ]
 doc_parts.defn_match_table(pids=examples)
 ```

docs/matching.md

@@ -1,6 +1,6 @@
 ---
 comment: |
-  ’’’ codebraid pandoc –katex –from markdown+tex_math_single_backslash –filter pandoc-sidenote  
+  ’’’ codebraid pandoc –katex –from markdown+tex_math_single_backslash –filter pandoc-sidenote
   –to html5+smart –template=$HOME/.pandoc/templates/template.html5 \
     --css=$HOME/.pandoc/theme.css –toc –wrap=none matching.md \> matching.html ’’’
 
@@ -49,7 +49,7 @@ The provided identifier string is split into several components (Figure 2) by ap
                   |    |             |
                scheme  |           value
                      prefix
-           
+
     scheme = "ark"
     content = "12345/some_value/with?extra=foo"
     prefix = "12345"

pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "rslv"
-version = "0.9.6"
+version = "0.10.0"
 description = "Provides an identifier resolver service in FastAPI."
 authors = [{ name = "datadavev", email = "605409+datadavev@users.noreply.github.com" }]
 requires-python = ">=3.9,<3.13"
@@ -28,6 +28,7 @@ dev = [
     "httpx>=0.28.1,<0.29",
     "flake8>=7.1.1,<8",
     "black>=25.1.0,<26",
+    "pre-commit>=4.3.0",
 ]
 cli = [
     "click>=8,<9",
@@ -47,4 +48,3 @@ default-groups = [
 ]
 
 [tool.poetry_bumpversion.file."rslv/__init__.py"]
-

requirements.txt

@@ -1,5 +1,5 @@
 # This file was autogenerated by uv via the following command:
-#    uv export --no-hashes --no-group dev --no-group docs --format requirements-txt
+#    uv export --no-hashes --no-dev --no-group docs --output-file=requirements.txt
 -e .
 annotated-types==0.7.0
     # via pydantic

rslv/config.py

@@ -46,6 +46,8 @@ class Settings(pydantic_settings.BaseSettings):
     # Note that this should be set False on services offering one-to-one matching of
     # definitions to PIDs. For N2T and arks.org this sould be true to match legacy behavior.
     auto_introspection: bool = True
+    # Optional header that if set, service returns a 200 code instead of redirect.
+    request_no_redirect: str = "x-no-redirect"
 
 
 def load_settings():

rslv/lib_rslv/__init__.py

@@ -39,14 +39,14 @@ def split_identifier_string(pid_str: str) -> typing.Dict[str, typing.Any]:
     parsed["scheme"] = _parts[0].strip().lower()
     try:
         parsed["content"] = _parts[1].lstrip(" /:")
-        parsed["content"] = parsed["content"].strip()  # type: ignore 
+        parsed["content"] = parsed["content"].strip()  # type: ignore
     except IndexError:
         return parsed
-    _parts = parsed["content"].split("/", 1)  # type: ignore 
+    _parts = parsed["content"].split("/", 1)  # type: ignore
     parsed["prefix"] = _parts[0].strip()
     try:
         parsed["value"] = _parts[1].lstrip(" /")
-        parsed["value"] = parsed["value"].strip() # type: ignore 
+        parsed["value"] = parsed["value"].strip() # type: ignore
     except IndexError:
         pass
     return parsed

rslv/lib_rslv/piddefine.py

@@ -442,12 +442,18 @@ def parse(
             )
             parts["suffix"] = pid_str[suffix_pos:]
 
-        # Hack alert - need to deal with the oddness of ARK identifiers ignoring hyphens.
+        # Hack alert - Optionally need to deal with the oddness of ARK identifiers ignoring hyphens.
         if parts["scheme"] == "ark":
-            # remove hyphens from the content and value portions, but not from the query portion, if present...
-            parts["content"] = rslv.lib_rslv.remove_hyphens(parts["content"])
-            parts["value"] = rslv.lib_rslv.remove_hyphens(parts["value"])
-            parts["suffix"] = rslv.lib_rslv.remove_hyphens(parts["suffix"])
+            # Hyphen stripping for ARKs is optionally set at the definition level
+            # and defaults to True to match the legacy resolver behavior
+            strip_ark_hyphens = True
+            if pid_definition.properties is not None:
+                strip_ark_hyphens = pid_definition.properties.get("strip_hyphens", True)
+            if strip_ark_hyphens:
+                # remove hyphens from the content and value portions, but not from the query portion, if present...
+                parts["content"] = rslv.lib_rslv.remove_hyphens(parts["content"])
+                parts["value"] = rslv.lib_rslv.remove_hyphens(parts["value"])
+                parts["suffix"] = rslv.lib_rslv.remove_hyphens(parts["suffix"])
         return parts, pid_definition
 
     def list_schemes(self, valid_targets_only: bool = False):

rslv/routers/resolver.py

@@ -122,12 +122,16 @@ def get_service_info(request: fastapi.Request, valid: bool = True):
 
 
 def handle_get_info(
-    request: fastapi.Request, cleaned_identifier: CleanedIdentifierRequest
+    request: fastapi.Request,
+    cleaned_identifier: CleanedIdentifierRequest,
+    pid_config,
+    pid_parts: dict,
+    definition: typing.Optional[rslv.lib_rslv.piddefine.PidDefinition]
 ):
-    pid_config = rslv.lib_rslv.piddefine.PidDefinitionCatalog(request.state.dbsession)
-    pid_parts, definition = pid_config.parse(
-        cleaned_identifier.cleaned, resolve_synonym=False
-    )
+    #pid_config = rslv.lib_rslv.piddefine.PidDefinitionCatalog(request.state.dbsession)
+    #pid_parts, definition = pid_config.parse(
+    #    cleaned_identifier.cleaned, resolve_synonym=False
+    #)
     # TODO: This is where a definition specific handler can be used for
     #   further processing of the PID, e.g. to remove hyphens from an ark.
     #   Basically, add a property to the definition that contains the name
@@ -216,7 +220,18 @@ def get_info(
         str(request.url), identifier, request.app.state.settings.service_pattern
     )
 
-    return handle_get_info(request, cleaned_identifier)
+    pid_config = rslv.lib_rslv.piddefine.PidDefinitionCatalog(request.state.dbsession)
+    pid_parts, definition = pid_config.parse(
+        cleaned_identifier.cleaned, resolve_synonym=False
+    )
+
+    return handle_get_info(
+        request,
+        cleaned_identifier,
+        pid_config,
+        pid_parts,
+        definition
+    )
 
 
 @router.head(
@@ -275,16 +290,22 @@ def get_resolve(
         str(request.url), identifier, request.app.state.settings.service_pattern
     )
 
-    # If the request was for introspection (inflection) use the info handler
-    if cleaned_identifier.is_introspection:
-        return handle_get_info(request, cleaned_identifier)
 
     # Get the identifier configuration catalog
     pid_config = rslv.lib_rslv.piddefine.PidDefinitionCatalog(request.state.dbsession)
 
     # Split the identifier string into components and find the best match from the catalog
     pid_parts, definition = pid_config.parse(cleaned_identifier.cleaned)
-    # TODO: see above in get_info for PID handling with specific schemes.
+
+    # If the request was for introspection (inflection) use the info handler
+    if cleaned_identifier.is_introspection:
+        return handle_get_info(
+            request,
+            cleaned_identifier,
+            pid_config,
+            pid_parts,
+            definition
+        )
 
     if definition is None:
         # Return a 404 response and include the pid parts in the body with a
@@ -307,7 +328,13 @@ def get_resolve(
         None,
         "",
     ]:
-        return handle_get_info(request, cleaned_identifier)
+        return handle_get_info(
+            request,
+            cleaned_identifier,
+            pid_config,
+            pid_parts,
+            definition
+        )
     # If the PID value part matches the value part of the matched definition,
     # then return the definition information. This is sketchy behavior but included
     # here because it follows the legacy N2T behavior. It can be disabled through
@@ -316,11 +343,21 @@ def get_resolve(
         request.app.state.settings.auto_introspection
         and pid_parts["value"] == definition.value
     ):
-        return handle_get_info(request, cleaned_identifier)
+        return handle_get_info(
+            request,
+            cleaned_identifier,
+            pid_config,
+            pid_parts,
+            definition
+        )
     # OK, past all the edge cases, redirect the client to the registered target.
     pid_parts["canonical"] = pid_format(pid_parts, definition.canonical)
     pid_parts["status_code"] = response_status_code
     headers = {"Location": _target}
+    # Check if request includes no redirect header and
+    # override the redirect if so.
+    if request.app.state.settings.request_no_redirect in request.headers:
+        response_status_code = 200
     return fastapi.responses.JSONResponse(
         content=pid_parts,
         headers=headers,

rslv/static/style.css

@@ -193,4 +193,4 @@ ul.blog-posts li a:visited {
     .helptext {
         color: #aaa;
     }
-}
+}

rslv/templates/index.html

@@ -34,4 +34,4 @@ <h1>rslv - Identifier Resolution Service</h1>
     </content>
 </main>
 <footer></footer>
-</body>
+</body>

tests/test_parser.py

@@ -71,6 +71,13 @@ def do_add(cfg, entry):
                 scheme="ark", prefix="12345", value="up", properties={"name": "Frank"}
             ),
         )
+        do_add(
+            cfg,
+            rslv.lib_rslv.piddefine.PidDefinition(
+                scheme="ark", prefix="12345", value="nostrip",
+                properties={"name": "Frank", "strip_hyphens":False}
+            ),
+        )
         cfg.refresh_metadata()
 
 
@@ -90,6 +97,10 @@ def do_add(cfg, entry):
         "bark:99999/fk44wlr;jglerig",
         {"scheme": "ark", "prefix": "99999", "value": "fk4"},
     ),
+    (
+        "ark:12345/nostrip-test",
+        {"scheme": "ark", "prefix": "12345", "value": "nostrip", "suffix":"-test"},
+    )
 )