Add Population tab and convert .csv to .xml (May need more adjust)

state.py

@@ -29,20 +29,6 @@ import utils
 verbosity = 4  # 0=None 1=Low 2=Medium 3=High 4=Debug
 
 def printV(message, level):
-    '''
-    Logging function to print based on verbosity level
-
-    ...
-
-    Parameters
-    ----------
-    message : str
-        Message to display
-    level : int
-        Target verbosity level of message; 0=None 1=Low 2=Medium 3=High 4=Debug
-    '''
-    # Read from global value above this function
-    global verbosity
     if verbosity >= level:
         print(message)
 
@@ -62,6 +48,7 @@ def init_session_state():
         "removeCategory":   None,
         "resetResult":      False,
         "unusedSpeakers":   {},   # {filename: [speaker, ...]}
+        "studentPopulations": {}, # {filename: int or None}
         "file_names":       [],
         "valid_files":      [],
         "file_paths":       {},   # {filename: path}
@@ -104,54 +91,96 @@ def convert_df(df):
     return df.to_csv(index=False).encode("utf-8")
 
 
-def build_all_csv_zip():
-    """Build an in-memory ZIP containing one CSV per analyzed file.
-
-    Applies the same transformations as the single-file download:
-    drop Task, rename Resource -> Speaker, sort by Start, add Role.
-    Returns raw ZIP bytes ready for st.download_button.
+def _build_analysis_df(fname):
+    """Build the cleaned analysis DataFrame for a single file (shared logic)."""
+    annotation, _ = st.session_state.results[fname]
+    currDF, _ = su.annotationToSimpleDataFrame(annotation)
+    raw_to_role = {
+        token.split(": ", 1)[1]: st.session_state.categories[i]
+        for i, tokens in enumerate(st.session_state.categorySelect)
+        for token in tokens
+        if token.startswith(f"{fname}: ")
+    }
+    currDF = currDF.copy()
+    currDF["Role"] = currDF["Resource"].map(raw_to_role).fillna("")
+    renames = st.session_state.speakerRenames.get(fname, {})
+    if "Resource" in currDF.columns:
+        currDF["Resource"] = currDF["Resource"].apply(lambda s: renames.get(s, s))
+    currDF = currDF.drop(columns=["Task"], errors="ignore")
+    currDF = currDF.rename(columns={"Resource": "Speaker"})
+    if "Start" in currDF.columns:
+        currDF = currDF.sort_values("Start").reset_index(drop=True)
+    return currDF
+
+
+def build_xml_download(fname):
+    """Build XML bytes for a single analyzed file.
+
+    Structure:
+        <recording filename="..." student_population="N">
+            <segment speaker="..." role="..." start="..." end="..."/>
+            ...
+        </recording>
     """
-    import io
-    import zipfile
-
+    import xml.etree.ElementTree as ET
+    df         = _build_analysis_df(fname)
+    population = st.session_state.studentPopulations.get(fname)
+    plain_name = fname.rsplit(".", 1)[0]
+
+    root = ET.Element("recording")
+    root.set("filename", plain_name)
+    root.set("student_population",
+             str(population) if population is not None else "")
+
+    for _, row in df.iterrows():
+        seg = ET.SubElement(root, "segment")
+        seg.set("speaker", str(row.get("Speaker", "")))
+        seg.set("role",    str(row.get("Role", "")))
+        seg.set("start",   str(row.get("Start", "")))
+        seg.set("end",     str(row.get("Finish", row.get("End", ""))))
+
+    tree = ET.ElementTree(root)
+    buf  = io.BytesIO()
+    ET.indent(tree, space="  ")
+    tree.write(buf, encoding="unicode", xml_declaration=True)
+    return buf.getvalue().encode("utf-8")
+
+
+def build_all_xml_zip():
+    """Build an in-memory ZIP containing one XML per analyzed file."""
     buf = io.BytesIO()
     with zipfile.ZipFile(buf, "w", zipfile.ZIP_DEFLATED) as zf:
         for fname, result in st.session_state.results.items():
             if len(result) != 2:
                 continue
             try:
-                annotation, _ = result
-                currDF, _ = su.annotationToSimpleDataFrame(annotation)
-
-                # Add Role column against raw SPEAKER_## labels BEFORE renames
-                raw_to_role = {
-                    token.split(": ", 1)[1]: st.session_state.categories[i]
-                    for i, tokens in enumerate(st.session_state.categorySelect)
-                    for token in tokens
-                    if token.startswith(f"{fname}: ")
-                }
-                currDF = currDF.copy()
-                currDF["Role"] = currDF["Resource"].map(raw_to_role).fillna("")
-
-                # Apply speaker renames after Role is set
-                renames = st.session_state.speakerRenames.get(fname, {})
-                if "Resource" in currDF.columns:
-                    currDF["Resource"] = currDF["Resource"].apply(
-                        lambda s: renames.get(s, s)
-                    )
-                currDF = currDF.drop(columns=["Task"], errors="ignore")
-                currDF = currDF.rename(columns={"Resource": "Speaker"})
-                if "Start" in currDF.columns:
-                    currDF = currDF.sort_values("Start").reset_index(drop=True)
+                plain_name = fname.rsplit(".", 1)[0]
+                zf.writestr(
+                    f"sonogram-analysis-{plain_name}.xml",
+                    build_xml_download(fname),
+                )
+            except Exception as e:
+                print(f"build_all_xml_zip: skipping {fname} — {e}")
+    buf.seek(0)
+    return buf.read()
+
 
+def build_all_csv_zip():
+    """Build an in-memory ZIP containing one CSV per analyzed file."""
+    buf = io.BytesIO()
+    with zipfile.ZipFile(buf, "w", zipfile.ZIP_DEFLATED) as zf:
+        for fname, result in st.session_state.results.items():
+            if len(result) != 2:
+                continue
+            try:
+                df         = _build_analysis_df(fname)
                 plain_name = fname.rsplit(".", 1)[0]
                 zf.writestr(
                     f"sonogram-analysis-{plain_name}.csv",
-                    currDF.to_csv(index=False)
+                    df.to_csv(index=False)
                 )
             except Exception as e:
                 print(f"build_all_csv_zip: skipping {fname} — {e}")
-
     buf.seek(0)
     return buf.read()
 
@@ -161,37 +190,19 @@ def build_all_csv_zip():
 # ---------------------------------------------------------------------------
 
 def addCategory():
-    '''
-    Create new category
-    '''
-    # Clean up input
     new = st.session_state.categoryInput.strip()
     if not new:
         return
-    # Notify user
     st.toast(f"Adding {new}")
-    # Update all relevant session_state variables
     st.session_state.categories.append(new)
     st.session_state.categorySelect.append([])
     st.session_state.pop(f"multiselect_{new}", None)
-    # Reset category input field
     st.session_state.categoryInput = ""
 
 
 def removeCategory(index):
-    '''
-    Remove category
-
-    Parameters
-    ----------
-    index : int
-        Index of category to be removed
-    '''
-    # Get name of category
     name = st.session_state.categories[index]
-    # Notify user
     st.toast(f"Removing {name}")
-    # Update all relevant session_state variables
     st.session_state.pop(f"multiselect_{name}", None)
     del st.session_state.categories[index]
     del st.session_state.categorySelect[index]
@@ -401,9 +412,6 @@ def on_grename_change(idx, token_display_map=None):
 # ---------------------------------------------------------------------------
 
 def updateMultiSelect():
-    '''
-    Updates multi-select field
-    '''
     fileName = st.session_state["select_currFile"]
     st.session_state.resetResult = True
     result = st.session_state.results.get(fileName)
@@ -645,21 +653,11 @@ def build_table_df(displayDF):
 # ---------------------------------------------------------------------------
 
 def analyze(inFileName):
-    """
-    Compute and store all summary DataFrames for inFileName.
-
-    Updates all session_state variables based on stored diarization results
-
-    Parameters
-    ----------
-    inFileName : str
-        Name of file to analyze
-    """
+    """Compute and store all summary DataFrames for inFileName."""
     try:
         printV(f"Start analyzing {inFileName}", 4)
         st.session_state.resetResult = False
 
-        # If file has not been processed by Sonogram, then end
         if not (
             inFileName in st.session_state.results
             and inFileName in st.session_state.summaries