handle more response types

.gitignore

@@ -1 +1,2 @@
 *.pyc
+dist/

src/whsk/__init__.py

@@ -41,10 +41,28 @@ def parse_headers(headers: list[str]) -> dict:
     return header_dict
 
 
-def make_request(url, headers, postdata):
+def make_request(url, *, headers, user_agent, postdata):
+    """
+    Helper function for redundant code between methods.
+
+    Will take all parameters related to request and return
+    a 2-tuple of the raw response and the parsed response.
+
+    The second parameter may be a:
+        - JSON dict
+        - lxml.html.HtmlElement
+        - lxml.etree.Element
+    """
     header_dict = parse_headers(headers)
-    resp = httpx.request("GET", url, headers=header_dict, data=postdata)
+    method = "GET"
-    # if resp.headers["content-type"] == "text/html":
+    if postdata:
+        method = "POST"
+    resp = httpx.request(method, url, headers=header_dict, data=postdata)
+    if resp.headers["content-type"] == "application/json":
+        root = resp.json()
+    elif "xml" in resp.headers["content-type"]:
+        root = lxml.etree.fromstring(resp.content)
+    else:
         root = lxml.html.fromstring(resp.text)
     return resp, root
 
@@ -95,9 +113,18 @@ def query(
     xpath: Annotated[str, opt["xpath"]] = "",
 ):
     """Run a one-off query against the URL"""
-    resp, root = make_request(url, headers, postdata)
+    resp, root = make_request(
+        url, headers=headers, user_agent=user_agent, postdata=postdata
+    )
+    if not isinstance(root, lxml.html.HtmlElement):
+        typer.secho(f"Expecting HTML response, got:\n{root}", fg="red")
+        raise typer.Exit(1)
     selector, selected = parse_selectors(root, css, xpath)
 
+    if selector is None:
+        typer.secho("Must provide either --css or --xpath to query", fg="red")
+        raise typer.Exit(1)
+
     for s in selected:
         print(s)
 
@@ -113,7 +140,9 @@ def shell(
 ):
     """Launch an interactive Python shell for scraping"""
 
-    resp, root = make_request(url, headers, postdata)
+    resp, root = make_request(
+        url, headers=headers, user_agent=user_agent, postdata=postdata
+    )
     selector, selected = parse_selectors(root, css, xpath)
 
     console = Console()
@@ -125,7 +154,13 @@ def shell(
     )
     table.add_row("[green]url[/green]", url)
     table.add_row("[green]resp[/green]", str(resp))
-    table.add_row("[green]root[/green]", "lxml.html.Element")
+    # if this list of parsed types expands there's probably a better way
+    if isinstance(root, lxml.html.HtmlElement):
+        table.add_row("[green]root[/green]", "lxml.html.HtmlElement")
+    elif isinstance(root, lxml.etree._Element):
+        table.add_row("[green]root[/green]", "lxml.etree.Element (XML)")
+    elif isinstance(root, dict):
+        table.add_row("[green]root[/green]", "dict (JSON)")
     if selector:
         table.add_row("[green]selector[/green]", selector)
         table.add_row("[green]selected[/green]", f"{len(selected)} elements")