Update state.py

state.py

@@ -247,6 +247,121 @@ def register_file(fname):
         st.session_state.file_names.append(fname)
 
 
+# ---------------------------------------------------------------------------
+# File loading helpers
+# ---------------------------------------------------------------------------
+
+def load_annotation_file(fname, fpath):
+    """Load an annotation-only file (.txt / .rttm / .csv) into session state."""
+    ext = fpath.lower()
+    if ext.endswith(".txt"):
+        _, annotations = su.loadAudioTXT(fpath)
+    elif ext.endswith(".rttm"):
+        _, annotations = su.loadAudioRTTM(fpath)
+    elif ext.endswith(".csv"):
+        _, annotations = su.loadAudioCSV(fpath)
+    else:
+        raise ValueError(f"Unsupported annotation format: {fpath}")
+    totalSeconds = max((s.end for s in annotations.itersegments()), default=0)
+    st.session_state.results[fname]        = (annotations, totalSeconds)
+    st.session_state.summaries[fname]      = {}
+    st.session_state.unusedSpeakers[fname] = list(annotations.labels())
+    return annotations, totalSeconds
+
+
+def load_demo_single(demo_path):
+    """Register and load a single RTTM demo file, then run analyze()."""
+    import time
+    dname = demo_path.split("/")[-1]
+    register_file(dname)
+    st.session_state.file_paths[dname] = demo_path
+    start_time = time.time()
+    with st.spinner("Loading Demo Sample"):
+        load_annotation_file(dname, demo_path)
+    with st.spinner("Analyzing Demo Data"):
+        analyze(dname)
+    st.success(f"Took {time.time() - start_time:.1f}s to analyze the demo file!")
+    st.session_state.select_currFile = dname
+    return dname
+
+
+def load_demo_multi(demo_paths):
+    """Register and load multiple RTTM demo files."""
+    for demo_path in demo_paths:
+        dname = demo_path.split("/")[-1]
+        register_file(dname)
+        st.session_state.file_paths[dname] = demo_path
+        with st.spinner(f"Loading: {dname}"):
+            load_annotation_file(dname, demo_path)
+    st.session_state.analyzeAllToggle = True
+
+
+def run_analysis_loop(file_names, file_paths_dict, pipeline,
+                      enable_denoise, early_cleanup,
+                      gain_window, minimum_gain, maximum_gain,
+                      df_model, df_state, atten_lim_db):
+    """Process every file in file_names and populate session state."""
+    import time
+    import utils as _utils
+    start_time = time.time()
+    totalFiles = len(file_names)
+
+    for i, fname in enumerate(file_names):
+        fpath = file_paths_dict.get(fname, "")
+        ext   = fpath.lower()
+
+        if ext.endswith((".txt", ".rttm", ".csv")):
+            label = ext.rsplit(".", 1)[-1].upper()
+            with st.spinner(f"Loading {label} {i+1}/{totalFiles}"):
+                load_annotation_file(fname, fpath)
+        else:
+            with st.spinner(f"Processing Audio {i+1}/{totalFiles}"):
+                annotations, totalSeconds, waveform, sample_rate = _utils.processFile(
+                    fpath, pipeline, enable_denoise, early_cleanup,
+                    gain_window, minimum_gain, maximum_gain,
+                    df_model, df_state, atten_lim_db,
+                )
+                st.session_state.results[fname]        = (annotations, totalSeconds)
+                st.session_state.summaries[fname]      = {}
+                st.session_state.unusedSpeakers[fname] = list(annotations.labels())
+            with st.spinner(f"Generating clips {i+1}/{totalFiles}"):
+                store_speaker_clips(fname, annotations, waveform, sample_rate)
+                del waveform
+
+        with st.spinner(f"Analyzing {i+1}/{totalFiles}"):
+            analyze(fname)
+
+    st.success(f"Analyzed {totalFiles} file(s) in {time.time() - start_time:.1f}s")
+    st.session_state.analyzeAllToggle = False
+
+
+def build_table_df(displayDF):
+    """Return a display-only copy of displayDF with cosmetic transforms applied:
+       - Rename 'Resource' -> 'Speaker'
+       - Drop 'Task' column if present
+       - Format Start / Finish as HH:MM:SS.cs strings
+    """
+    def _fmt(val):
+        try:
+            secs = float(val)
+        except (TypeError, ValueError):
+            return str(val)
+        h  = int(secs // 3600)
+        m  = int(secs % 3600 // 60)
+        s  = int(secs % 60)
+        cs = round((secs % 1) * 100)
+        return f"{h:02d}:{m:02d}:{s:02d}.{cs:02d}"
+
+    df = displayDF.copy()
+    if "Task" in df.columns:
+        df = df.drop(columns=["Task"])
+    if "Start" in df.columns:
+        df["Start"] = df["Start"].apply(_fmt)
+    if "Finish" in df.columns:
+        df["Finish"] = df["Finish"].apply(_fmt)
+    return df.rename(columns={"Resource": "Speaker"})
+
+
 # ---------------------------------------------------------------------------
 # analyze() — build and cache all DataFrames for one file
 # ---------------------------------------------------------------------------