atsfill 0.0.5.dev181__tar.gz → 0.0.5.dev182__tar.gz

{atsfill-0.0.5.dev181 → atsfill-0.0.5.dev182}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: atsfill
-Version: 0.0.5.dev181
+Version: 0.0.5.dev182
 Summary: Read, dry-run and submit job application forms on hiring systems: Ashby, Greenhouse, Teamtailor, Welcome to the Jungle.
 Keywords: job-applications,ats,ashby,greenhouse,teamtailor,ai-agents,cli
 Author: shbernal

{atsfill-0.0.5.dev181 → atsfill-0.0.5.dev182}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "atsfill"
-version = "0.0.5.dev181"
+version = "0.0.5.dev182"
 description = "Read, dry-run and submit job application forms on hiring systems: Ashby, Greenhouse, Teamtailor, Welcome to the Jungle."
 readme = "README.md"
 license = "MIT"

{atsfill-0.0.5.dev181 → atsfill-0.0.5.dev182}/pyproject.toml.orig

@@ -1,6 +1,6 @@
 [project]
 name = "atsfill"
-version = "0.0.5.dev181"
+version = "0.0.5.dev182"
 description = "Read, dry-run and submit job application forms on hiring systems: Ashby, Greenhouse, Teamtailor, Welcome to the Jungle."
 readme = "README.md"
 license = "MIT"

{atsfill-0.0.5.dev181 → atsfill-0.0.5.dev182}/src/atsfill/adapters/ashby.py

@@ -78,7 +78,9 @@ class Ashby:
 
     def __init__(self) -> None:
         self._forms: list[dict[str, Any]] = []  # the application form, then each survey
-        self._pending: set[Any] = set()  # GraphQL requests in flight, since `browser_read`
+        self._pending: set[Any] = set()  # GraphQL requests in flight, since `open_form`
+        self._listening = False
+        self._loaded: Response | None = None  # the ApiJobPosting response `open_form` caught
         self._last_change = 0.0
         self._hosted: str | None = None  # the jobs.ashbyhq.com posting a custom domain embeds
 
@@ -125,14 +127,24 @@ class Ashby:
         except OSError, ValueError, KeyError, TypeError, Unsupported:
             return None
 
-    def browser_read(self, page: Page) -> list[Field]:
-        page.on("request", self._on_request)
-        page.on("requestfinished", self._on_finished)
-        page.on("requestfailed", self._on_finished)
-        page.on("response", self._on_response)
+    def open_form(self, page: Page, url: str) -> None:
+        """Loads `/application`, listening first for the page's own ApiJobPosting response."""
+        if not self._listening:
+            page.on("request", self._on_request)
+            page.on("requestfinished", self._on_finished)
+            page.on("requestfailed", self._on_finished)
+            page.on("response", self._on_response)
+            self._listening = True
         with page.expect_response(lambda r: _op(r) == "ApiJobPosting") as posting:
-            page.goto(self.posting(page.url) + "/application")
-        job = posting.value.json()["data"]["jobPosting"]
+            page.goto(self.posting(url) + "/application")
+        self._loaded = posting.value
+
+    def browser_read(self, page: Page) -> list[Field]:
+        if self._loaded is None:  # a page claimed once loaded: an embed
+            self.open_form(page, page.url)
+        assert self._loaded
+        loaded, self._loaded = self._loaded, None
+        job = loaded.json()["data"]["jobPosting"]
         if job is None:
             raise Unsupported("posting_gone")
         self._forms = [job["applicationForm"], *job["surveyForms"]]

{atsfill-0.0.5.dev181 → atsfill-0.0.5.dev182}/src/atsfill/adapters/breezy.py

@@ -211,6 +211,10 @@ _FILE_SHOWN = """e => {
 }"""
 
 
+def _on_form(url: str) -> bool:
+    return urlsplit(url).path.rstrip("/").endswith("/apply")
+
+
 class Breezy:
     name = "breezy"
     known_gaps: dict[str, Gap] = {}
@@ -236,9 +240,12 @@ class Breezy:
         match = JOB.match(parts.path)
         return f"https://{parts.hostname}/p/{match[1]}" if match else url
 
+    def open_form(self, page: Page, url: str) -> None:
+        page.goto(url if _on_form(url) else f"{self.posting(url)}/apply")
+
     def browser_read(self, page: Page) -> list[Field]:
-        if not urlsplit(page.url).path.rstrip("/").endswith("/apply"):
-            page.goto(f"{self.posting(page.url)}/apply")
+        if not _on_form(page.url):
+            self.open_form(page, page.url)
         form = page.locator(FORM)
         try:
             form.locator("[name=cName]").wait_for(state="attached", timeout=FORM_TIMEOUT_MS)

{atsfill-0.0.5.dev181 → atsfill-0.0.5.dev182}/src/atsfill/adapters/lever.py

@@ -116,6 +116,10 @@ _READ_BACK = """(form, kinds) => {
 }"""
 
 
+def _on_form(url: str) -> bool:
+    return urlsplit(url).path.rstrip("/").endswith("/apply")
+
+
 class Lever:
     name = "lever"
     known_gaps: dict[str, Gap] = {}
@@ -134,9 +138,12 @@ class Lever:
         parts, job = urlsplit(url), _job(url)
         return f"https://{parts.hostname}/{job[0]}/{job[1]}" if job else url
 
+    def open_form(self, page: Page, url: str) -> None:
+        page.goto(url if _on_form(url) else f"{self.posting(url)}/apply")
+
     def browser_read(self, page: Page) -> list[Field]:
-        if not urlsplit(page.url).path.rstrip("/").endswith("/apply"):
-            page.goto(f"{self.posting(page.url)}/apply")
+        if not _on_form(page.url):
+            self.open_form(page, page.url)
         form = page.locator(FORM)
         try:
             form.wait_for(state="attached", timeout=FORM_TIMEOUT_MS)

{atsfill-0.0.5.dev181 → atsfill-0.0.5.dev182}/src/atsfill/adapters/manatal.py

@@ -12,7 +12,7 @@ from typing import Any
 from urllib.parse import urlsplit
 
 from atsfill.core.adapter import Blocked, Gap, Unsupported, known_gap
-from atsfill.core.browser import HTTP_TIMEOUT_S, Page
+from atsfill.core.browser import HTTP_TIMEOUT_S, Page, Response
 from atsfill.core.results import Answer, Field, FieldType, Is, SideEffect
 
 HOST = "www.careers-page.com"
@@ -183,6 +183,7 @@ class Manatal:
 
     def __init__(self) -> None:
         self._entries: list[dict[str, Any]] = []
+        self._loaded: Response | None = None  # the form's GET `open_form` caught
 
     def detect(self, url: str, page: Page | None) -> bool:
         parts = urlsplit(url).path.strip("/").split("/")
@@ -210,13 +211,20 @@ class Manatal:
         except OSError, ValueError, KeyError, TypeError:
             return None
 
-    def browser_read(self, page: Page) -> list[Field]:
+    def open_form(self, page: Page, url: str) -> None:
+        """Loads `/apply`, listening first for the form's GET."""
         with page.expect_response(
             lambda r: r.request.method == "GET" and "/application-form/" in r.url,
             timeout=LOAD_TIMEOUT_MS,
         ) as loaded:
-            page.goto(_posting(page.url) + "/apply")
-        response = loaded.value
+            page.goto(_posting(url) + "/apply")
+        self._loaded = loaded.value
+
+    def browser_read(self, page: Page) -> list[Field]:
+        if self._loaded is None:
+            self.open_form(page, page.url)
+        assert self._loaded
+        response, self._loaded = self._loaded, None
         if response.headers.get("x-amzn-waf-action") == "challenge":
             raise Blocked("captcha: AWS WAF challenge")
         if not response.ok:

{atsfill-0.0.5.dev181 → atsfill-0.0.5.dev182}/src/atsfill/adapters/recruitee.py

@@ -62,6 +62,10 @@ _READ_BACK = """(form, kinds) => {
 }"""
 
 
+def _on_form(url: str) -> bool:
+    return urlsplit(url).path.rstrip("/").endswith("/c/new")
+
+
 class Recruitee:
     name = "recruitee"
     known_gaps: dict[str, Gap] = {}
@@ -84,9 +88,12 @@ class Recruitee:
         match = JOB.match(parts.path)
         return f"https://{parts.hostname}/o/{match[1]}" if match else url
 
+    def open_form(self, page: Page, url: str) -> None:
+        page.goto(url if _on_form(url) else f"{self.posting(url)}/c/new")
+
     def browser_read(self, page: Page) -> list[Field]:
-        if not urlsplit(page.url).path.rstrip("/").endswith("/c/new"):
-            page.goto(f"{self.posting(page.url)}/c/new")
+        if not _on_form(page.url):
+            self.open_form(page, page.url)
         form = page.locator(FORM)
         try:
             form.wait_for(state="attached", timeout=FORM_TIMEOUT_MS)

{atsfill-0.0.5.dev181 → atsfill-0.0.5.dev182}/src/atsfill/adapters/teamtailor.py

@@ -125,6 +125,7 @@ class Teamtailor:
         self._follow_ups: set[str] = set()
         self._uploads: dict[str, str] = {}  # file field id: the answer's file name
         self._videos: dict[str, tuple[str, str]] = {}  # video field id: its video id, its URL
+        self._form: str | None = None  # the form `open_form` loaded
 
     def detect(self, url: str, page: Page | None) -> bool:
         """A `*.teamtailor.com` job, or a job page with Teamtailor's "powered by" link or its
@@ -141,13 +142,21 @@ class Teamtailor:
         assert match, url
         return f"https://{parts.hostname}/jobs/{match[1]}"
 
-    def browser_read(self, page: Page) -> list[Field]:
+    def open_form(self, page: Page, url: str) -> None:
         """Opens the form's own page, never the overlay nor "Apply with LinkedIn"."""
-        parts = urlsplit(page.url)
+        parts = urlsplit(url)
         job = JOB.search(parts.path)
-        assert job, page.url  # detected on a job URL
-        page.goto(f"{parts.scheme}://{parts.netloc}{parts.path[: job.end()]}/applications/new")
-        if urlsplit(page.url).hostname != parts.hostname:
+        assert job, url  # detected on a job URL
+        form = f"{parts.scheme}://{parts.netloc}{parts.path[: job.end()]}/applications/new"
+        page.goto(form)
+        self._form = form
+
+    def browser_read(self, page: Page) -> list[Field]:
+        if self._form is None:  # a page claimed once loaded: a custom domain
+            self.open_form(page, page.url)
+        assert self._form
+        opened, self._form = self._form, None
+        if urlsplit(page.url).hostname != urlsplit(opened).hostname:
             raise Unsupported(f"external_step: the job applies on another site ({page.url})")
         form = page.locator(FORM)
         try:

{atsfill-0.0.5.dev181 → atsfill-0.0.5.dev182}/src/atsfill/adapters/workable.py

@@ -141,6 +141,10 @@ _READ_BACK = """(form, kinds) => {
 }"""
 
 
+def _on_form(url: str) -> bool:
+    return "/apply" in urlsplit(url).path
+
+
 class Workable:
     name = "workable"
     known_gaps = {"entries": Gap.RENDERING}
@@ -174,9 +178,12 @@ class Workable:
         except OSError, ValueError, KeyError, TypeError:
             return None
 
+    def open_form(self, page: Page, url: str) -> None:
+        page.goto(url if _on_form(url) else f"{self.posting(url)}/apply/")
+
     def browser_read(self, page: Page) -> list[Field]:
-        if "/apply" not in urlsplit(page.url).path:
-            page.goto(f"{self.posting(page.url)}/apply/")
+        if not _on_form(page.url):
+            self.open_form(page, page.url)
         form = page.locator(FORM)
         try:
             form.wait_for(timeout=FORM_TIMEOUT_MS)

{atsfill-0.0.5.dev181 → atsfill-0.0.5.dev182}/src/atsfill/core/adapter.py

@@ -280,6 +280,16 @@ class SendsToOpen(Adapter, Protocol):
         ...
 
 
+@runtime_checkable
+class OpensForm(Adapter, Protocol):
+    """An adapter whose form is another page than the posting: the core has it load the form
+    in place of the posting, when it claims the posting by its URL (ADR 0003)."""
+
+    def open_form(self, page: Page, url: str) -> None:
+        """Loads the form of the posting at `url`, so that `browser_read` finds it open."""
+        ...
+
+
 @runtime_checkable
 class ReadsByRequest(Adapter, Protocol):
     """An adapter with a request read: a form's fields listed without a browser (ADR 0011)."""

{atsfill-0.0.5.dev181 → atsfill-0.0.5.dev182}/src/atsfill/core/commands.py

@@ -29,6 +29,7 @@ from atsfill.core.adapter import (
     ListsShownFields,
     NeedsAccount,
     NotTaken,
+    OpensForm,
     ProfileIncomplete,
     ReadsByRequest,
     Refused,
@@ -173,7 +174,7 @@ def _opened(
             session = None
     with open_page(data, session, seconds) as page:
         try:
-            page.goto(url)
+            _load(page, adapter, url)
         except browser.Timeout:
             if adapter is None:
                 raise
@@ -207,6 +208,14 @@ def _opened(
                     browser.save_session(page, session)
 
 
+def _load(page: Page, adapter: object, url: str) -> None:
+    """Loads the posting, or its form where the adapter that claims it opens it (ADR 0003)."""
+    if isinstance(adapter, OpensForm):
+        adapter.open_form(page, url)
+    else:
+        page.goto(url)
+
+
 _again: ContextVar[Callable[[], Result | None] | None] = ContextVar("again", default=None)
 
 
@@ -281,7 +290,7 @@ def _sign_in_again(
             return _login_required(url, adapter, user)
         session = accounts.session_file(mine, adapter.name, scope)
         accounts.sign_in(page, adapter, scope, *credentials, session)
-        page.goto(url)
+        _load(page, adapter, url)
         return None
     email = credentials[0] if credentials else ""
     scoped = url if adapter.per_employer else None