Add codespell support (config, workflow to detect/not fix) and make it fix few typos

.codespellrc

@@ -0,0 +1,6 @@
+[codespell]
+# Ref: https://github.com/codespell-project/codespell#using-a-config-file
+skip = .git*,*.pdf,*.svg,.codespellrc
+check-hidden = true
+# ignore-regex = 
+# ignore-words-list =

.github/workflows/codespell.yml

@@ -0,0 +1,25 @@
+# Codespell configuration is within .codespellrc
+---
+name: Codespell
+
+on:
+  push:
+    branches: [main]
+  pull_request:
+    branches: [main]
+
+permissions:
+  contents: read
+
+jobs:
+  codespell:
+    name: Check for spelling errors
+    runs-on: ubuntu-latest
+
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v4
+      - name: Annotate locations with typos
+        uses: codespell-project/codespell-problem-matcher@v1
+      - name: Codespell
+        uses: codespell-project/actions-codespell@v2

.pre-commit-config.yaml

@@ -7,4 +7,9 @@ repos:
         entry: black
         args: ["knowledge_storm/"]
         language: python
-        pass_filenames: true
+        pass_filenames: true
+  - repo: https://github.com/codespell-project/codespell
+    # Configuration for codespell is in .codespellrc
+    rev: v2.3.0
+    hooks:
+    - id: codespell

examples/storm_examples/run_storm_wiki_serper.py

@@ -77,7 +77,7 @@ def main(args):
     # Important to note that tbs(date range is hardcoded values).
     # num is results per pages and is recommended to use in increments of 10(10, 20, etc).
     # page is how many pages will be searched.
-    # h1 is where the google search will orginate from.
+    # h1 is where the google search will originate from.
     topic = input("topic: ")
     data = {"autocorrect": True, "num": 10, "page": 1}
     rm = SerperRM(serper_search_api_key=os.getenv("SERPER_API_KEY"), query_params=data)

frontend/demo_light/pages_util/CreateNewArticle.py

@@ -81,7 +81,7 @@ def handle_final_writing():
 
             # update status bar
             st.session_state["page3_write_article_state"] = "prepare_to_show_result"
-            status.update(label="information snythesis complete!", state="complete")
+            status.update(label="information synthesis complete!", state="complete")
 
 def handle_prepare_to_show_result():
     if st.session_state["page3_write_article_state"] == "prepare_to_show_result":

knowledge_storm/collaborative_storm/modules/costorm_expert_utterance_generator.py

@@ -33,7 +33,7 @@ class GenExpertActionPlanning(dspy.Signature):
     last_utterance = dspy.InputField(
         prefix="Last utterance in the conversation: \n", format=str
     )
-    resposne = dspy.OutputField(
+    response = dspy.OutputField(
         prefix="Now give your note. Start with one of [Original Question, Further Details, Information Request, Potential Answer] with one sentence description\n",
         format=str,
     )
@@ -126,7 +126,7 @@ def forward(
                         expert=current_expert,
                         summary=conversation_summary,
                         last_utterance=last_utterance,
-                    ).resposne
+                    ).response
                 action_type, action_content = self.parse_action(action)
 
         if self.callback_handler is not None:

knowledge_storm/collaborative_storm/modules/information_insertion_module.py

@@ -79,7 +79,7 @@ def _get_navigation_choice(
         if child_names:
             structure += f"Child Nodes: {child_names}"
         navigated_path = " -> ".join(knowledge_node.get_path_from_root())
-        structure += f"Path you have nagivated: {navigated_path}"
+        structure += f"Path you have navigated: {navigated_path}"
 
         # get predicted action
         with dspy.settings.context(lm=self.engine):

knowledge_storm/lm.py

@@ -488,7 +488,7 @@ def __call__(self, prompt, only_completed=True, return_sorted=False, **kwargs):
         assert return_sorted is False, "for now"
         # per eg here: https://docs.anthropic.com/claude/reference/messages-examples
         # max tokens can be used as a proxy to return smaller responses
-        # so this cannot be a proper indicator for incomplete response unless it isnt the user-intent.
+        # so this cannot be a proper indicator for incomplete response unless it isn't the user-intent.
         n = kwargs.pop("n", 1)
         completions = []
         for _ in range(n):

knowledge_storm/storm_wiki/modules/storm_dataclass.py

@@ -387,7 +387,7 @@ def pre_order_find_index(node):
                     pre_order_find_index(child)
 
         pre_order_find_index(self.root)
-        # constrcut index mapping
+        # construct index mapping
         ref_index_mapping = {}
         for ref_index in ref_indices:
             if ref_index not in ref_index_mapping: