Add command line options

cmaureir · cmaureir · commit 7bd76cba77a5 · 2025-07-21T21:28:07.000+02:00
diff --git a/README.md b/README.md
@@ -49,3 +49,26 @@ The fields for speakers are the following:
 * Title: Title of the proposal
 * Level: If provided by the schedule. If two levels are displayed, pick the
     most general one
+
+## Scripts
+
+> [!NOTE]
+> Make sure to install the dependencies in `scripts/requirements.txt`
+
+In case the event you are trying to contribute is using Pretalx, you can get
+the speaker list by running the command:
+
+```
+python scripts/get_speakers.py --url https://pretalx.com/api/events/pyladiescon-2025
+```
+To write the output directly into a file, you can use the `--output` option,
+and execute the script like this:
+
+```
+python scripts/get_speakers.py \
+    --url https://pretalx.com/api/events/pyladiescon-2024 \
+    --output data/speakers/PyLadiesCon/2024.json
+```
+
+> [!IMPORTANT]
+> Make sure that the URL needs to the the one pointing to the API.
diff --git a/scripts/get_speakers.py b/scripts/get_speakers.py
@@ -1,114 +1,141 @@
 import json
+import sys
+import argparse
+from typing import Any
+from datetime import datetime
+
 import requests
 
-# Configuration variables:
-# This URL can be a custom URL or pretalx.com/ one depending on the event configuration
-CONFERENCE_URL = ""
-TOKEN = ""  # Not necessary, but in case of private events it's required
-YEAR = 2025  # Some events will not have it in the URL so we do it manually
 
+def get_response_data_from_url(url: str) -> dict[str, Any]:
+    try:
+        response = requests.get(url, headers={"Authorization": TOKEN})
+        data = response.json()
+    except requests.exceptions.RequestException as e:
+        print(f"Error fetching data: {e}")
+        sys.exit(1)
 
-def get_pretalx_submission_types():
-    url = f"{CONFERENCE_URL}/submission-types"
-    submission_types = {}
+    return data
 
-    while url:
-        try:
-            response = requests.get(url, headers={"Authorization": TOKEN})
-            data = response.json()
 
-            for stype in data["results"]:
-                if stype["id"] not in submission_types:
-                    submission_types[stype["id"]] = stype["name"]
+def get_pretalx_submission_types(conference_url: str) -> dict[str, str]:
+    url: str | None = f"{conference_url}/submission-types"
+    submission_types: dict[str, str] = {}
+
+    while url:
+        data: dict[str, Any] = get_response_data_from_url(url)
 
-            url = data.get("next")
+        for stype in data["results"]:
+            if stype["id"] not in submission_types:
+                submission_types[stype["id"]] = stype["name"]
 
-        except requests.exceptions.RequestException as e:
-            print(f"Error fetching submission types: {e}")
-            return None
+        url = data.get("next")
 
     return submission_types
 
 
-def get_pretalx_submissions(all_speakers):
-    url = f"{CONFERENCE_URL}/submissions"
+def get_pretalx_submissions(conference_url: str, all_speakers: dict[str, str]) -> dict[str, dict]:
+    url: str | None = f"{conference_url}/submissions"
     confirmed_submissions = {}
 
-    submissions_types = get_pretalx_submission_types()
+    submissions_types = get_pretalx_submission_types(conference_url)
 
     while url:
-        try:
-            response = requests.get(url, headers={"Authorization": TOKEN})
-
-            data = response.json()
-
-            for submission in data["results"]:
+        data: dict[str, Any] = get_response_data_from_url(url)
 
-                if submission["state"] == "confirmed":
+        for submission in data["results"]:
 
-                    submission_code = submission["code"]
-                    submission_speakers = submission["speakers"]
-                    submission_type = submissions_types[submission["submission_type"]]
-                    if isinstance(submission_type, dict):
-                        try:
-                            # Search for 'en' default language
-                            # otherwise, the first one
-                            submission_type = submission_type["en"]
-                        except KeyError:
-                            for k, v in submission_type.items():
-                                submission_type = v
-                                break
+            if submission["state"] == "confirmed":
 
-                    for speaker_code in submission["speakers"]:
+                submission_code = submission["code"]
+                submission_speakers = submission["speakers"]
+                submission_type = submissions_types[submission["submission_type"]]
+                if isinstance(submission_type, dict):
+                    try:
+                        # Search for 'en' default language
+                        # otherwise, the first one
+                        submission_type = submission_type["en"]
+                    except KeyError:
+                        for k, v in submission_type.items():
+                            submission_type = v
+                            break
 
-                        speaker_name = all_speakers[speaker_code]
-                        confirmed_submissions[f"{submission_code}-{speaker_code}"] = {
-                            "fullname": speaker_name,
-                            "title": submission["title"],
-                            "type": submission_type,
-                        }
+                for speaker_code in submission_speakers:
 
-            url = data.get("next")
+                    speaker_name = all_speakers[speaker_code]
+                    confirmed_submissions[f"{submission_code}-{speaker_code}"] = {
+                        "fullname": speaker_name,
+                        "title": submission["title"],
+                        "type": submission_type,
+                    }
 
-        except requests.exceptions.RequestException as e:
-            print(f"Error fetching submissions: {e}")
-            return None
+        url = data.get("next")
 
     return confirmed_submissions
 
 
-def get_pretalx_speakers():
-    url = f"{CONFERENCE_URL}/speakers"
+def get_pretalx_speakers(conference_url: str) -> dict[str, str]:
+    url: str | None = f"{conference_url}/speakers"
     speakers = {}
 
     while url:
-        try:
-            response = requests.get(url, headers={"Authorization": TOKEN})
-            data = response.json()
+        data: dict[str, Any] = get_response_data_from_url(url)
 
-            for speaker in data["results"]:
-                code = speaker["code"]
-                if code not in speakers:
-                    speakers[code] = speaker["name"]
+        for speaker in data["results"]:
+            code = speaker["code"]
+            if code not in speakers:
+                speakers[code] = speaker["name"]
 
-            url = data.get("next")
-
-        except requests.exceptions.RequestException as e:
-            print(f"Error fetching speakers: {e}")
-            return None
+        url = data.get("next")
 
     return speakers
 
 
-if __name__ == "__main__":
-    speakers = get_pretalx_speakers()
+def get_event_year(conference_url: str) -> int:
+    data: dict[str, Any] = get_response_data_from_url(conference_url)
+    date = data["date_from"]
+    year: int = 2025
+    try:
+        date_from = datetime.strptime(date, "%Y-%m-%d")
+    except ValueError:
+        print(f"Error: couldn't parse date '{date}'. Falling back to 2025")
+        return year
+
+    return date_from.year
 
-    submission_data = get_pretalx_submissions(speakers)
-    data = {}
+
+if __name__ == "__main__":
+    # This URL can be a custom URL or pretalx.com/ one depending on the event configuration
+    CONFERENCE_URL = ""
+    TOKEN = ""  # Not necessary, but in case of private events it's required
+
+    parser = argparse.ArgumentParser()
+    parser.add_argument('--url')
+    parser.add_argument('--output')
+    args = parser.parse_args()
+
+    # Check command-line arguments
+    if args.url is not None:
+        if CONFERENCE_URL:
+            print(f"-- WARNING: Overriding url '{CONFERENCE_URL}' "
+                  f"with command line value '{args.url}'")
+        print(f"-- Using event url: '{args.url}'")
+        CONFERENCE_URL = args.url
+
+    if args.output is not None:
+        if not args.output.endswith(".json"):
+            print(f"ERROR: file '{args.output}' is not a JSON file")
+            sys.exit(1)
+
+    year = get_event_year(CONFERENCE_URL)
+    speakers = get_pretalx_speakers(CONFERENCE_URL)
+    submission_data = get_pretalx_submissions(CONFERENCE_URL, speakers)
+
+    data: dict[str, Any] = {}
 
     if submission_data:
 
-        data["year"] = YEAR
+        data["year"] = year
         data["speakers"] = []
         for _, entry in submission_data.items():
             data["speakers"].append(
@@ -119,7 +146,11 @@ def get_pretalx_speakers():
                 }
             )
 
-        print(json.dumps(data, indent=4))
-
+        if args.output is not None:
+            with open(args.output, "w") as f:
+                json.dump(data, f, ensure_ascii=False, indent=4)
+            print(f"-- Written file: '{args.output}'")
+        else:
+            print(json.dumps(data, indent=4))
     else:
         print("Failed to fetch speakers data")
diff --git a/scripts/requirements.txt b/scripts/requirements.txt
@@ -0,0 +1,2 @@
+requests
+types-requests
