visidata 2.11.1__py3-none-any.whl → 3.0__py3-none-any.whl

visidata/loaders/mbtiles.py

@@ -5,11 +5,11 @@ import sqlite3
 
 @VisiData.api
 def open_pbf(vd, p):
-    return PbfSheet(p.name, source=p)
+    return PbfSheet(p.base_stem, source=p)
 
 @VisiData.api
 def open_mbtiles(vd, p):
-    return MbtilesSheet(p.name, source=p)
+    return MbtilesSheet(p.base_stem, source=p)
 
 def getListDepth(L):
     if not isinstance(L, list):
@@ -36,7 +36,7 @@ class MbtilesSheet(Sheet):
     ]
 
     def getTile(self, zoom_level, tile_col, tile_row):
-        import mapbox_vector_tile
+        mapbox_vector_tile = vd.importExternal('mapbox_vector_tile', 'mapbox-vector-tile')
 
         con = sqlite3.connect(str(self.source))
         tile_data = con.execute('''
@@ -128,11 +128,10 @@ class PbfCanvas(InvertedCanvas):
                     if disptext:
                         self.label(textx, texty, disptext, attr, row)
 
-
         self.refresh()
 
 
 PbfSheet.addCommand('.', 'plot-row', 'vd.push(PbfCanvas(name+"_map", source=sheet, sourceRows=[cursorRow], textCol=cursorCol))', 'plot blocks in current row')
 PbfSheet.addCommand('g.', 'plot-rows', 'vd.push(PbfCanvas(name+"_map", source=sheet, sourceRows=rows, textCol=cursorCol))', 'plot selected blocks')
 MbtilesSheet.addCommand('.', 'plot-row', 'vd.push(getPlot(cursorRow))', 'plot tiles in current row')
-MbtilesSheet.addCommand('g.', 'plot-selected', 'vd.push(getPlot(*selectedRows))', 'plot selected tiles'),
+MbtilesSheet.addCommand('g.', 'plot-selected', 'vd.push(getPlot(*selectedRows))', 'plot selected tiles')

visidata/loaders/mysql.py

@@ -1,6 +1,7 @@
 from contextlib import contextmanager
+from urllib.parse import urlparse, unquote
 
-from visidata import VisiData, vd, Sheet, anytype, asyncthread, urlparse, ColumnItem
+from visidata import VisiData, vd, Sheet, anytype, asyncthread, ColumnItem
 
 def codeToType(type_code, colname):
     import MySQLdb as mysql
@@ -41,7 +42,7 @@ class SQL:
                     database=self.url.path[1:],
                     host=self.url.hostname,
                     port=self.url.port or 3306,
-                    password=self.url.password,
+                    password=unquote(self.url.password),
                     use_unicode=True,
                     charset='utf8',
                     cursorclass=cursors.SSCursor) ## if SSCursor is not used mysql will first fetch ALL data, and only then visualize it
@@ -71,7 +72,7 @@ def cursorToColumns(cur, sheet):
 class MyTablesSheet(Sheet):
     rowtype = 'tables'
 
-    def reload(self):
+    def iterload(self):
         qstr = f'''
             select
                 t.table_name,
@@ -96,16 +97,15 @@ class MyTablesSheet(Sheet):
         '''
 
         with self.sql.cur(qstr) as cur:
-            self.rows = []
             # try to get first row to make cur.description available
             r = cur.fetchone()
             if r:
-                self.addRow(r)
+                yield r
             cursorToColumns(cur, self)
             self.setKeys(self.columns[0:1])  # table_name is the key
 
             for r in cur:
-                self.addRow(r)
+                yield r
 
     def openRow(self, row):
         return MyTable(self.name+"."+row[0], source=row[0], sql=self.sql)
@@ -113,22 +113,20 @@ class MyTablesSheet(Sheet):
 
 # rowdef: tuple of values as returned by fetchone()
 class MyTable(Sheet):
-    @asyncthread
-    def reload(self):
+    def iterload(self):
         with self.sql.cur("SELECT * FROM " + self.source) as cur:
-            self.rows = []
             r = cur.fetchone()
             if r is None:
                 return
-                
-            self.addRow(r)
+
+            yield r
             cursorToColumns(cur, self)
             while True:
                 try:
                     r = cur.fetchone()
                     if r is None:
                         break
-                        
-                    self.addRow(r)
+
+                    yield r
                 except UnicodeDecodeError as e:
                     vd.exceptionCaught(e)

visidata/loaders/npy.py

@@ -1,20 +1,20 @@
-from visidata import VisiData, vd, Sheet, date, anytype, options, Column, Progress, ColumnItem, vlen, PyobjSheet, TypedWrapper, InferColumnsSheet
+from visidata import VisiData, vd, Sheet, date, anytype, options, Column, Progress, ColumnItem, vlen, PyobjSheet, TypedWrapper
 
 'Loaders for .npy and .npz.  Save to .npy.  Depends on the zip loader.'
 
 @VisiData.api
 def open_npy(vd, p):
-    return NpySheet(p.name, source=p)
+    return NpySheet(p.base_stem, source=p)
 
 @VisiData.api
 def open_npz(vd, p):
-    return NpzSheet(p.name, source=p)
+    return NpzSheet(p.base_stem, source=p)
 
 vd.option('npy_allow_pickle', False, 'numpy allow unpickling objects (unsafe)')
 
 class NpySheet(Sheet):
     def iterload(self):
-        import numpy
+        numpy = vd.importExternal('numpy')
         if not hasattr(self, 'npy'):
             self.npy = numpy.load(str(self.source), encoding='bytes', **self.options.getall('npy_'))
         self.reloadCols()
@@ -47,12 +47,12 @@ class NpzSheet(vd.ZipSheet):
     ]
 
     def iterload(self):
-        import numpy
+        numpy = vd.importExternal('numpy')
         self.npz = numpy.load(str(self.source), encoding='bytes', **self.options.getall('npy_'))
         yield from Progress(self.npz.items())
 
     def openRow(self, row):
-        import numpy
+        numpy = vd.importExternal('numpy')
         tablename, tbl = row
         if isinstance(tbl, numpy.ndarray):
             return NpySheet(tablename, npy=tbl)
@@ -62,7 +62,7 @@ class NpzSheet(vd.ZipSheet):
 
 @VisiData.api
 def save_npy(vd, p, sheet):
-    import numpy as np
+    np = vd.importExternal('numpy')
 
     dtype = []
 

visidata/loaders/odf.py

@@ -3,11 +3,12 @@ from visidata import vd, VisiData, Sheet, IndexSheet, SequenceSheet
 
 @VisiData.api
 def open_ods(vd, p):
-    return OdsIndexSheet(p.name, source=p)
+    return OdsIndexSheet(p.base_stem, source=p)
 
 
 class OdsIndexSheet(IndexSheet):
     def iterload(self):
+        vd.importExternal('odf', 'odfpy')
         import odf.opendocument
         import odf.table
         self.doc = odf.opendocument.load(self.source)
@@ -16,6 +17,7 @@ class OdsIndexSheet(IndexSheet):
 
 
 def _get_cell_string_value(cell, text_s):
+        vd.importExternal('odf', 'odfpy')
         from odf.element import Element
         from odf.namespaces import TEXTNS
 
@@ -34,6 +36,7 @@ def _get_cell_string_value(cell, text_s):
 
 class OdsSheet(SequenceSheet):
     def iterload(self):
+        vd.importExternal('odf', 'odfpy')
         import odf.table
         import odf.text
         from odf.namespaces import TABLENS, OFFICENS

visidata/loaders/orgmode.py

@@ -0,0 +1,428 @@
+'''
+View sections in one or more .org files.  Easily edit sections within the system editor.
+
+## Supported syntax
+
+- `--- [comment]` at start of line starts a new section
+- any number of `#` or `*` (followed by a space) leads a headline
+- `#+key: value` adds metadata to next element (headline, table, section, )
+- orgmode style tags: `:tag1:tag2:`
+- `[[tagname]]` links to that set of tags
+- `[[url]]` links to external url
+- orgmode and markdown links
+   - `[[url][linktext]]`
+   - `[linktext](url)`
+- all other markup/orgmode/whatever is ignored and passed through
+
+## Usage
+
+- `vd file.org`
+- or `find orgfiles/ -name '*.org' | vd -f forg`
+- or `vd orgfiles/ -f orgdir`
+'''
+
+import collections
+import datetime
+import os
+import re
+
+from visidata import vd, VisiData, Column, Sheet, ItemColumn, vlen, asyncthread, Path, AttrDict, date
+
+
+@VisiData.api
+def open_org(vd, p):
+    return OrgSheet(p.base_stem, source=p, filetype='org')
+
+
+@VisiData.api
+def open_forg(vd, p):
+    return OrgSheet(p.base_stem, source=p, filetype='forg')
+
+
+@VisiData.api
+def open_orgdir(vd, p):
+    return OrgSheet(p.base_stem, source=p, filetype='orgdir')
+
+
+def encode_date(dt=None):
+    if not dt:
+        dt = datetime.datetime.now()
+    elif isinstance(dt, str):
+        dt = datetime.datetime.fromisoformat(dt)
+
+    s = '123456789abcdefghijklmnopqrstuvwxyz'
+    return '%02d%s%s' % (dt.year % 100, s[dt.month-1], s[dt.day-1])
+
+
+class OrgContentsColumn(Column):
+    def setValue(self, row, v):
+        super().setValue(row, v)
+        orgmode_parse_into(row, v)
+
+    def putValue(self, row, v):
+        self.sheet.save(row)
+
+
+def sectionize(lines):
+    'Generate (startinglinenum, contentlines) for each section.  First section may not have leading * or # but all others will.'
+    startinglinenum = 0
+    prev_contents = contents = []
+    for linenum, line in enumerate(lines):
+        line = line.rstrip('\n')
+
+        if line and line[0] in '#*':
+            if contents:
+                yield startinglinenum, contents
+                prev_contents = contents
+                contents = []
+                startinglinenum = linenum
+        else:
+            if not line and not contents:
+                prev_contents.append(line)
+
+        contents.append(line)
+
+    if contents:
+        yield startinglinenum, contents
+
+
+def orgmode_parse(all_lines):
+    root = parent = OrgSheet().newRow()
+    for linenum, lines in sectionize(all_lines):
+        section = OrgSheet().newRow()
+
+        section.contents = ''
+        for i, line in enumerate(lines):
+            if not line and not lines[i-1]:
+                continue
+            section.contents += line + '\n'
+        section.orig_contents = section.contents
+
+        section.linenum = linenum+1
+        for line in lines:
+            section.tags.extend(re.findall(r':([\S:]+?):', line))
+            links = re.findall(r'\[.*?\]\(.*?\)', line)
+            if links:
+                section.links.extend(links)
+
+        title = orgmode_parse_title(lines[0])
+        if not title:
+            root = parent = section
+            continue
+
+        section.update(title)
+
+        while parent and section.level <= parent.level:
+            parent = parent.parent
+
+        parent.children.append(section)
+        section.parent = parent
+        parent = section
+
+    return root
+
+def _replace(node, newnode):
+    node.update(newnode)  # must replace insides of same row object
+    for c in node.children:
+        c.parent = node
+    return node
+
+def orgmode_parse_into(toprow, text):
+    row = orgmode_parse(text.splitlines())
+    if not row.title and len(row.children) == 1:
+        row = row.children[0]
+        row.parent = toprow.parent
+    toprow = _replace(toprow, row)
+
+    return toprow
+
+
+def orgmode_to_string(section, prestars=''):
+    ret = ''
+
+#    if section.title:
+#        ret += prestars[len(section.stars):] + (section.title or ' ') + '\n'
+    ret += section.contents.rstrip() or ''
+    ret += '\n\n'
+    ret += ''.join(orgmode_to_string(c, prestars+(section.stars or '')) for c in section.children).rstrip() + '\n'
+
+    return ret
+
+
+def orgmode_parse_title(line):
+    m = re.match(r'^(?P<stars>[*#]+)\s*(?P<keyword>(TODO|FEEDBACK|VERIFY|DONE|DELEGATED))?\s*(?P<prio>\[#[A-z]\])?\s*(?P<title>.*)', line)
+    if not m:
+        assert not line or line[0] not in '#*', line
+        return {}
+
+    return dict(stars=m.group('stars'),
+                level=len(m.group('stars')),
+                keyword=m.group('keyword') or '',
+                prio=m.group('prio') or '',
+                title=line)
+
+
+class OrgSheet(Sheet):
+    guide = '''# Orgmode Sheet (experimental)
+A list of orgmode sections from _{sheet.source}_.
+
+- `Enter` to expand current section
+- `z Enter` to contract current section
+- `ga` to combine selected sections into a new entry
+- `Ctrl+O` to edit section in system editor (edits source directly)
+- `g Ctrl+S` to save all org files
+'''
+    defer = True
+    columns = [
+        Column('path', getter=lambda c,r: _root(r).path, width=0),
+        Column('id', getter=lambda c,r: _root(r).path.stem),
+        ItemColumn('title', width=40),
+#        ItemColumn('date', width=0, type=date),
+        ItemColumn('tags', width=10, type=lambda v: ' '.join(v)),
+        ItemColumn('links', type=vlen),
+#        ItemColumn('parent', width=0),
+        ItemColumn('children', type=vlen),
+        ItemColumn('linenum', width=0, type=int),
+        Column('to_string', width=0, getter=lambda c,r: orgmode_to_string(r)),
+        OrgContentsColumn('orig_contents', width=0, getter=lambda c,r: r.orig_contents),
+    ]
+
+    def __init__(self, *args, **kwargs):
+        super().__init__(*args, **kwargs)
+        self.opened_rows = set()
+#        self.open_max=0  # all levels closed after this point
+#        self.close_max=0 # all levels open after this point
+
+    def isSelectedParents(self, row):
+        return super().isSelected(row) or row.parent and self.isSelectedParents(row.parent)
+
+    def isSelected(self, row):
+        return self.isSelectedParents(row)
+
+    def refreshRows(self):
+        self.rows = list(self._deepiter(self.sourceRows))
+
+    def _deepiter(self, objlist, depth=1):
+        for obj in objlist:
+            if not obj.parent and obj.children:
+                # ignore toplevel file, dive into subrows
+                yield from self._deepiter(obj.children, depth)
+            else:
+                yield obj
+                if id(obj) in self.opened_rows:
+                    yield from self._deepiter(obj.children, depth-1)
+
+    def openRows(self, rows):
+        for row in rows:
+            self.opened_rows.add(id(row))
+        self.refreshRows()
+
+    def closeRows(self, rows=None):
+        if rows is None:
+            for row in rows:
+                self.opened_rows.remove(id(row))
+        else:
+            self.opened_rows.clear()
+        self.refreshRows()
+
+    def newRow(self):
+        return AttrDict(title='', contents='', tags=[], children=[], links=[], level=0, linenum=0)
+
+    def iterload(self):
+        self.rows = []
+        def _walkfiles(p):
+            basepath = str(p)
+            for folder, subdirs, files in os.walk(basepath):
+                subfolder = folder[len(basepath)+1:]
+                if subfolder.startswith('.'): continue
+                if subfolder in ['.', '..']: continue
+
+                fpath = Path(folder)
+                yield fpath
+
+                for fn in files:
+                    yield fpath/fn
+
+        if self.filetype == 'orgdir':
+            basepath = str(self.source)
+            for p in _walkfiles(self.source):
+                if p.base_stem.startswith('.'): continue
+                if p.ext in ['org', 'md']:
+                    yield self.parse_orgmd(p)
+        elif self.filetype == 'forg':
+            for fn in self.source.open():
+                yield self.parse_orgmd(Path(fn.rstrip()))
+        elif self.filetype == 'org':
+            yield self.parse_orgmd(self.source)
+
+        self.sourceRows = self.rows
+        self.refreshRows()
+
+    def parse_orgmd(self, path):
+#        row.file_string = open(path).read()
+        row = orgmode_parse(open(path).readlines())
+        st = path.stat()
+        if st:
+            mtime = st.st_mtime
+
+        row.path = path
+        row.date = mtime
+        return row
+
+    def draw(self, scr):
+        super().draw(scr)
+
+    @asyncthread
+    def putChanges(self):
+        adds, mods, dels = self.getDeferredChanges()
+
+        saveset = {}  # path:_root(row)
+        for r in adds.values():
+            saveset[_root(r).path] = _root(r)
+        for r, _ in mods.values():
+            saveset[_root(r).path] = _root(r)
+
+        for row in addset.values():
+            self.save(row)
+
+        self.commitAdds()
+
+        self.commitMods()
+
+        for row in dels.values():
+            row.path.rename('%s-%s' % (row.path, encode_date()))
+
+        self.commitDeletes()
+
+    def save_all(self):
+        for row in self.sourceRows:
+            self.save(row)
+        vd.status('saved %s org files' % len(self.sourceRows))
+
+    def save(self, row):
+        try:
+            row.path.rename('%s-%s' % (row.path, encode_date()))  # backup
+        except FileNotFoundError:
+            pass
+
+        with row.path.open(mode='w') as f:
+            print(orgmode_to_string(row).rstrip(), file=f)
+
+
+@OrgSheet.api
+def paste_into(sheet, row, sourcerows, cols):
+    row.children.extend(sourcerows)
+    for r in sourcerows:
+        r.parent.children.remove(r)
+        r.parent = row
+    sheet.refreshRows()
+
+
+@OrgSheet.api
+def paste_data_into(sheet, row, sourcerows, cols):
+    body = row.body or ''
+    for r in sourcerows:
+        data = vd.encode_json(r, cols)
+        row.contents += f':{cols[0].sheet.name}:{data}\n'
+
+
+@OrgSheet.api
+def combine_rows(sheet, rows):
+    newrow = sheet.newRow()
+    newrow.date = datetime.datetime.today()
+    orgid = vd.cleanName(rows[0].title)
+    newrow.path = Path((orgid or encode_date())+'.org')
+    for r in rows:
+        hdr = sheet.newRow()
+        if hdr.title:
+            hdr.update(orgmode_parse_title(hdr.title))
+        hdr.children = list(r.children)
+        newrow.children.append(hdr)
+#    newrow.title = ' '
+    newrow.file_string = orgmode_to_string(newrow, '*')
+    return newrow
+
+
+def _root(row):
+    while row and row.parent:
+        row = row.parent
+    return row
+
+
+@OrgSheet.api
+def sysopen_row(sheet, row):
+    root = _root(row)
+    if root.path.exists():
+        vd.launchEditor(root.path, '+%s'%row['linenum'])
+    else:
+        orgmode_parse_into(row, vd.launchExternalEditor(root.file_string))
+
+@VisiData.api
+def save_org(vd, p, *vsheets):
+    with p.open(mode='w', encoding=vsheets[0].options.save_encoding) as fp:
+        for vs in vsheets:
+            if isinstance(vs, OrgSheet):
+                for row in vs.rows:
+                    print(orgmode_to_string(row).strip(), file=fp)
+            else:
+                vd.warning('not implemented')
+
+
+@OrgSheet.api
+def sysopen_rows(sheet, rows):
+    ret = ''
+    for r in rows:
+        s = orgmode_to_string(r).strip()
+        ret += '{::%s::}\n%s\n\n' % (id(r), s)
+
+    ret = vd.launchExternalEditor(ret + "{::}")
+
+    idrows = {id(r):r for r in rows}
+
+    idtexts = {}
+    for idtag, text in re.findall(r'\{::(\d+)::\}(.*?)(?={::)', ret, re.DOTALL):
+        idtexts[int(idtag)] = text
+
+    # reparse all existing
+    for rowid, row in idrows.items():
+        text = idtexts.get(rowid, '').strip()
+        if text:
+            orgmode_parse_into(row, text)
+        else:
+            # sometimes, top levels are missing parents
+            # this needs to be debugged
+            if row.parent:
+                row.parent.children.remove(row)
+
+    lastrow = None
+    # find new rows to add
+    for rowid, text in idtexts.items():
+        if rowid not in idrows:
+            section = orgmode_parse(text.splitlines())
+            while section.level <= lastrow.level:
+                lastrow = lastrow.parent
+
+            if lastrow:
+                sourceRows.append(section)
+            else:
+                sheet.addRow(section)
+
+    sheet.refreshRows()
+
+OrgSheet.addCommand('^O', 'sysopen-row', 'sysopen_row(cursorRow)', 'open current file in external $EDITOR')
+OrgSheet.addCommand('g^O', 'sysopen-rows', 'sysopen_rows(selectedRows)', 'open selected files in external $EDITOR')
+OrgSheet.addCommand('^J', 'expand-row', 'openRows([cursorRow]); sheet.cursorRowIndex += 1')
+OrgSheet.addCommand('z^J', 'close-row', 'closeRows([cursorRow]); sheet.cursorRowIndex += 1')
+OrgSheet.addCommand('g^J', 'expand-selected', 'openRows(selectedRows)')
+OrgSheet.addCommand('gz^J', 'close-selected', 'closeRows(selectedRows)')
+OrgSheet.addCommand('ga', 'combine-selected', 'addRows([combine_rows(selectedRows)], index=cursorRowIndex); cursorDown(1)', 'combine selected rows into new org entry')
+
+OrgSheet.addCommand('zp', 'paste-data', 'paste_data_into(cursorRow, vd.memory.cliprows, vd.memory.clipcols)', 'move clipboard rows to children of current row')
+OrgSheet.addCommand('p', 'paste-sections', 'paste_data_into(cursorRow, vd.memory.cliprows, vd.memory.clipcols)', 'move clipboard rows to children of current row')
+OrgSheet.addCommand('g^S', 'save-all', 'save_all()', 'save all org files')
+
+
+if __name__ == '__main__':
+    for fn in sys.argv[1:]:
+        sect = orgmode_parse(Path(fn))
+        print(orgmode_to_string(sect).strip())

visidata/loaders/pandas_freqtbl.py

@@ -1,4 +1,4 @@
-from visidata import vd, Sheet, options, Column, asyncthread, Progress, PivotGroupRow, ENTER
+from visidata import vd, Sheet, options, Column, asyncthread, Progress, PivotGroupRow, ENTER, HistogramColumn
 
 from visidata.loaders._pandas import PandasSheet
 from visidata.pivot import PivotSheet
@@ -11,8 +11,8 @@ class DataFrameRowSliceAdapter:
     by visidata's selectRow implementation.
     """
     def __init__(self, df, mask):
-        import pandas as pd
-        import numpy as np
+        pd = vd.importExternal('pandas')
+        np = vd.importExternal('numpy')
         if not isinstance(df, pd.DataFrame):
             vd.fail('%s is not a dataframe' % type(df).__name__)
         if not isinstance(mask, pd.Series):
@@ -62,15 +62,15 @@ class DataFrameRowSliceIter:
         self.index += 1
         return row
 
+def makePandasFreqTable(sheet, *groupByCols):
+    fqcolname = '%s_freq' % '-'.join(col.name for col in groupByCols)
+    return PandasFreqTableSheet(sheet.name, fqcolname, groupByCols=groupByCols, source=sheet)
+
+
 class PandasFreqTableSheet(PivotSheet):
     'Generate frequency-table sheet on currently selected column.'
     rowtype = 'bins'  # rowdef FreqRow(keys, sourcerows)
 
-    def __init__(self, sheet, *groupByCols):
-        fqcolname = '%s_%s_freq' % (sheet.name, '-'.join(col.name for col in groupByCols))
-        super().__init__(fqcolname, groupByCols, [], source=sheet)
-        self.largest = 1
-
     def selectRow(self, row):
         # Select all entries in the bin on the source sheet.
         # Use the internally defined _selectByLoc to avoid
@@ -86,8 +86,7 @@ class PandasFreqTableSheet(PivotSheet):
     def updateLargest(self, grouprow):
         self.largest = max(self.largest, len(grouprow.sourcerows))
 
-    @asyncthread
-    def reload(self):
+    def loader(self):
         'Generate frequency table then reverse-sort by length.'
         import pandas as pd
 
@@ -95,8 +94,6 @@ class PandasFreqTableSheet(PivotSheet):
         # (e.g. as a histogram). We currently don't provide support for this
         # for PandasSheet, although we could implement it with a pd.Grouper
         # that operates similarly to pd.cut.
-        super().initCols()
-
         df = self.source.df.copy()
 
         # Implementation (special case): for one row, this degenerates
@@ -141,9 +138,7 @@ class PandasFreqTableSheet(PivotSheet):
                            getter=lambda col,row: len(row.sourcerows)),
                     Column('percent', type=float,
                            getter=lambda col,row: len(row.sourcerows)*100/df.shape[0]),
-                    Column('histogram', type=str,
-                           getter=lambda col,row: options.disp_histogram*(options.disp_histolen*len(row.sourcerows)//value_counts.max()),
-                           width=options.disp_histolen+2),
+                    HistogramColumn('histogram', type=str, width=self.options.default_width*2)
                     ]:
             self.addColumn(c)
 
@@ -174,11 +169,10 @@ def expand_source_rows(sheet, row):
         vd.fail("no source rows")
     return PandasSheet(sheet.name, vd.valueNames(row.discrete_keys, row.numeric_key), source=row.sourcerows)
 
-PandasSheet.addCommand('F', 'freq-col', 'vd.push(PandasFreqTableSheet(sheet, cursorCol))', 'open Frequency Table grouped on current column, with aggregations of other columns')
-PandasSheet.addCommand('gF', 'freq-keys', 'vd.push(PandasFreqTableSheet(sheet, *keyCols))', 'open Frequency Table grouped by all key columns on source sheet, with aggregations of other columns')
+PandasSheet.addCommand('F', 'freq-col', 'vd.push(makePandasFreqTable(sheet, cursorCol))', 'open Frequency Table grouped on current column, with aggregations of other columns')
+PandasSheet.addCommand('gF', 'freq-keys', 'vd.push(makePandasFreqTable(sheet, *keyCols))', 'open Frequency Table grouped by all key columns on source sheet, with aggregations of other columns')
 
+PandasFreqTableSheet.init('largest', lambda: 1)
 PandasFreqTableSheet.options.numeric_binning = False
 
-vd.addGlobals({
-        'PandasFreqTableSheet': PandasFreqTableSheet,
-        })
+vd.addGlobals(makePandasFreqTable=makePandasFreqTable)