visidata 2.11.dev0__py3-none-any.whl → 3.0__py3-none-any.whl

visidata/indexsheet.py

@@ -0,0 +1,109 @@
+from visidata import vd, VisiData, BaseSheet, Sheet, Column, AttrColumn, ItemColumn, setitem, asyncthread
+
+
+class IndexSheet(Sheet):
+    'Base class for tabular sheets with rows that are Sheets.'
+    guide = '''
+    # Index Sheet
+    This is a list of sheets from `{sheet.source}`.
+
+    - `Enter` to open {sheet.cursorRow}
+    - `g Enter` to open all selected sheets
+    '''
+    rowtype = 'sheets'  # rowdef: Sheet
+
+    columns = [
+        Column('name', getter=lambda c,r: r.names[-1], setter=lambda c,r,v: setitem(r.names, -1, v)),
+        AttrColumn('rows', 'nRows', type=int, width=9),
+        AttrColumn('cols', 'nCols', type=int),
+        AttrColumn('keys', 'keyColNames'),
+        AttrColumn('source'),
+    ]
+    nKeys = 1
+
+    def newRow(self):
+        return Sheet('', columns=[ItemColumn('', 0)], rows=[])
+
+    def openRow(self, row):
+        return row  # rowdef is Sheet
+
+    def getSheet(self, k):
+        for vs in self.rows:
+            if vs.name == k:
+                return vs
+
+    def addRow(self, sheet, **kwargs):
+        super().addRow(sheet, **kwargs)
+        if not self.options.load_lazy and not sheet.options.load_lazy:
+            sheet.ensureLoaded()
+
+    @asyncthread
+    def reloadSheets(self, sheets):
+        for vs in vd.Progress(sheets):
+            vs.reload()
+
+
+class SheetsSheet(IndexSheet):
+    columns = [
+        AttrColumn('name'),
+        AttrColumn('type', '__class__.__name__'),
+        AttrColumn('pane', type=int),
+        Column('shortcut', getter=lambda c,r: getattr(r, 'shortcut'), setter=lambda c,r,v: setattr(r, '_shortcut', v)),
+        AttrColumn('nRows', type=int),
+        AttrColumn('nCols', type=int),
+        AttrColumn('nVisibleCols', type=int),
+        AttrColumn('cursorDisplay'),
+        AttrColumn('keyColNames'),
+        AttrColumn('source'),
+        AttrColumn('progressPct'),
+#        AttrColumn('threads', 'currentThreads', type=vlen),
+    ]
+    precious = False
+    nKeys = 1
+    def reload(self):
+        self.rows = self.source
+
+    def sort(self):
+        self.rows[1:] = sorted(self.rows[1:], key=self.sortkey)
+
+
+class GlobalSheetsSheet(SheetsSheet):  #1620
+    def sort(self):
+        IndexSheet.sort(self)
+
+
+@VisiData.lazy_property
+def sheetsSheet(vd):
+    return SheetsSheet("sheets", source=vd.sheets)
+
+
+@VisiData.lazy_property
+def allSheetsSheet(vd):
+    return GlobalSheetsSheet("sheets_all", source=vd.allSheets)
+
+
+
+@Sheet.api
+def nextRow(sheet, n=1):
+    sheet.cursorRowIndex += n
+    sheet.checkCursor()
+    return sheet.rows[sheet.cursorRowIndex]  # cursorRow itself might be cached
+
+
+vd.addCommand('S', 'sheets-stack', 'vd.push(vd.sheetsSheet)', 'open Sheets Stack: join or jump between the active sheets on the current stack')
+vd.addCommand('gS', 'sheets-all', 'vd.push(vd.allSheetsSheet)', 'open Sheets Sheet: join or jump between all sheets from current session')
+
+BaseSheet.addCommand('g>', 'open-source-next', 'vd.replace(openSource(source.nextRow())) if isinstance(source, IndexSheet) else fail("parent sheet must be Index Sheet")', 'open next sheet on parent index sheet')
+BaseSheet.addCommand('g<', 'open-source-prev', 'vd.replace(openSource(source.nextRow(-1))) if isinstance(source, IndexSheet) else fail("parent sheet must be Index Sheet")', 'open prev sheet on parent index sheet')
+
+IndexSheet.addCommand('g^R', 'reload-selected', 'reloadSheets(selectedRows or rows)', 'reload all selected sheets')
+
+# when diving into a sheet, remove the index unless it is precious
+SheetsSheet.addCommand('gC', 'columns-selected', 'vd.push(ColumnsSheet("all_columns", source=selectedRows))', 'open Columns Sheet with all visible columns from selected sheets')
+SheetsSheet.addCommand('z^C', 'cancel-row', 'cancelThread(*cursorRow.currentThreads)', 'abort async thread for current sheet')
+SheetsSheet.addCommand('gz^C', 'cancel-rows', 'for vs in selectedRows: cancelThread(*vs.currentThreads)', 'abort async threads for selected sheets')
+SheetsSheet.addCommand('Enter', 'open-row', 'dest=cursorRow; vd.sheets.remove(sheet) if not sheet.precious else None; vd.push(openRow(dest))', 'open sheet referenced in current row')
+
+vd.addGlobals(IndexSheet=IndexSheet,
+              SheetsSheet=SheetsSheet,
+              GlobalSheetsSheet=GlobalSheetsSheet)

visidata/input_history.py

@@ -0,0 +1,55 @@
+from visidata import vd, VisiData, Sheet, ItemColumn, asyncthread
+
+
+vd._inputHistoryList = vd.StoredList(name='input_history')
+vd.inputHistory = {}   # [input_type][input] -> anything
+
+
+@VisiData.api
+def addInputHistory(vd, input:str, type:str=''):
+    r = vd.processInputHistory(input, type)
+    if r:
+        vd._inputHistoryList.append(r)
+
+
+@VisiData.api
+def processInputHistory(vd, input:str, type:str=''):
+    hist = list(vd.inputHistory.setdefault(type, {}).keys())
+    if hist and hist[-1] == input:
+        return
+    if input in vd.inputHistory[type]:
+        n = vd.inputHistory[type][input].get('n', 0)
+        del vd.inputHistory[type][input]  # make it the most recent entry
+    else:
+        n = 0
+
+    r = dict(type=type, input=input, n=n+1)
+    vd.inputHistory[type][input] = r
+    return r
+
+
+class InputHistorySheet(Sheet):
+    # rowdef: dict(type=, input=, n=)
+    # .source=vd.inputHistory
+    columns = [
+        ItemColumn('type'),
+        ItemColumn('input'),
+    ]
+    def iterload(self):
+        yield from vd._inputHistoryList
+
+
+@VisiData.before
+@asyncthread
+def run(vd, *args, **kwargs):
+    vd._inputHistoryList.reload()
+    for x in vd._inputHistoryList:
+        vd.processInputHistory(x.input, x.type)
+
+
+@VisiData.property
+def inputHistorySheet(vd):
+    return InputHistorySheet('input_history', source=vd._inputHistoryList.path)
+
+
+vd.addCommand(None, 'open-input-history', 'vd.push(inputHistorySheet)', 'open sheet with previous inputs')

visidata/interface.py

@@ -0,0 +1,58 @@
+from visidata import VisiData, vd
+
+vd.theme_option('disp_splitwin_pct', 0, 'height of second sheet on screen')
+vd.theme_option('disp_note_none', '⌀',  'visible contents of a cell whose value is None')
+vd.theme_option('disp_truncator', '…', 'indicator that the contents are only partially visible')
+vd.theme_option('disp_oddspace', '\u00b7', 'displayable character for odd whitespace')
+vd.theme_option('disp_more_left', '<', 'header note indicating more columns to the left')
+vd.theme_option('disp_more_right', '>', 'header note indicating more columns to the right')
+vd.theme_option('disp_error_val', '', 'displayed contents for computation exception')
+vd.theme_option('disp_ambig_width', 1, 'width to use for unicode chars marked ambiguous')
+
+vd.theme_option('disp_pending', '', 'string to display in pending cells')
+vd.theme_option('note_pending', '⌛', 'note to display for pending cells')
+vd.theme_option('note_format_exc', '?', 'cell note for an exception during formatting')
+vd.theme_option('note_getter_exc', '!', 'cell note for an exception during computation')
+vd.theme_option('note_type_exc', '!', 'cell note for an exception during type conversion')
+
+vd.theme_option('color_note_pending', 'bold magenta', 'color of note in pending cells')
+vd.theme_option('color_note_type', '226 yellow', 'color of cell note for non-str types in anytype columns')
+vd.theme_option('color_note_row', '220 yellow', 'color of row note on left edge')
+vd.option('scroll_incr', -3, 'amount to scroll with scrollwheel')
+vd.theme_option('disp_column_sep', '│', 'separator between columns')
+vd.theme_option('disp_keycol_sep', '║', 'separator between key columns and rest of columns')
+vd.theme_option('disp_rowtop_sep', '│', '') # ╷│┬╽⌜⌐▇
+vd.theme_option('disp_rowmid_sep', '⁝', '') # ┃┊│█
+vd.theme_option('disp_rowbot_sep', '⁝', '') # ┊┴╿⌞█⍿╵⎢┴⌊  ⋮⁝
+vd.theme_option('disp_rowend_sep', '║', '') # ┊┴╿⌞█⍿╵⎢┴⌊
+vd.theme_option('disp_keytop_sep', '║', '') # ╽╿┃╖╟
+vd.theme_option('disp_keymid_sep', '║', '') # ╽╿┃
+vd.theme_option('disp_keybot_sep', '║', '') # ╽╿┃╜‖
+vd.theme_option('disp_endtop_sep', '║', '') # ╽╿┃╖╢
+vd.theme_option('disp_endmid_sep', '║', '') # ╽╿┃
+vd.theme_option('disp_endbot_sep', '║', '') # ╽╿┃╜‖
+vd.theme_option('disp_selected_note', '•', '') #
+vd.theme_option('disp_sort_asc', '↑↟⇞⇡⇧⇑', 'characters for ascending sort') # ↑▲↟↥↾↿⇞⇡⇧⇈⤉⤒⥔⥘⥜⥠⍏˄ˆ
+vd.theme_option('disp_sort_desc', '↓↡⇟⇣⇩⇓', 'characters for descending sort') # ↓▼↡↧⇂⇃⇟⇣⇩⇊⤈⤓⥕⥙⥝⥡⍖˅ˇ
+vd.theme_option('color_default', 'white on black', 'the default fg and bg colors')
+vd.theme_option('color_default_hdr', 'bold', 'color of the column headers')
+vd.theme_option('color_bottom_hdr', 'underline', 'color of the bottom header row')
+vd.theme_option('color_current_row', 'reverse', 'color of the cursor row')
+vd.theme_option('color_current_col', 'bold', 'color of the cursor column')
+vd.theme_option('color_current_cell', '', 'color of current cell, if different from color_current_row+color_current_col')
+vd.theme_option('color_current_hdr', 'bold reverse', 'color of the header for the cursor column')
+vd.theme_option('color_column_sep', '246 blue', 'color of column separators')
+vd.theme_option('color_key_col', '81 cyan', 'color of key columns')
+vd.theme_option('color_hidden_col', '8', 'color of hidden columns on metasheets')
+vd.theme_option('color_selected_row', '215 yellow', 'color of selected rows')
+vd.theme_option('color_clickable', 'underline', 'color of internally clickable item')
+vd.theme_option('color_code', 'bold white on 237', 'color of code sample')
+vd.theme_option('color_heading', 'bold 200', 'color of header')
+vd.theme_option('color_guide_unwritten', '243 on black', 'color of unwritten guides in GuideGuide')
+
+vd.theme_option('force_256_colors', False, 'use 256 colors even if curses reports fewer')
+
+vd.option('quitguard', False, 'confirm before quitting modified sheet')
+vd.option('default_width', 20, 'default column width', replay=True, max_help=1)   # TODO: make not replay and remove from markdown saver
+vd.option('default_height', 4, 'default column height', max_help=-1)
+vd.option('textwrap_cells', True, 'wordwrap text for multiline rows', max_help=1)

visidata/keys.py

@@ -8,6 +8,8 @@ visidata.vd.prettykeys_trdict = {
         '^J': 'Enter',
         '^M': 'Enter',
         '^I': 'Tab',
+        'KEY_BTAB': 'Shift+Tab',
+        '^@': 'Ctrl+Space',
         'KEY_UP':    'Up',
         'KEY_DOWN':  'Down',
         'KEY_LEFT':  'Left',
@@ -30,6 +32,8 @@ visidata.vd.prettykeys_trdict = {
         'kNXT5':     'Ctrl+PgDn',
         'KEY_IC5':   'Ctrl+Ins',
         'KEY_DC5':   'Ctrl+Del',
+        'kDC5':      'Ctrl+Del',
+        'KEY_SDC':   'Shift+Del',
 
         'KEY_IC':    'Ins',
         'KEY_DC':    'Del',
@@ -47,24 +51,21 @@ visidata.vd.prettykeys_trdict = {
         'BUTTON1_PRESSED': 'LeftClick',
         'BUTTON2_PRESSED': 'MiddleClick',
         'BUTTON3_PRESSED': 'RightClick',
-        'BUTTON4_PRESSED': 'ScrollwheelUp',
-        'BUTTON5_PRESSED': 'ScrollwheelDown',
-        'REPORT_MOUSE_POSITION': 'ScrollwheelDown',
-        '2097152': 'ScrollwheelDown',
-        'KEY_F(1)': 'F1',
-        'KEY_F(2)': 'F2',
-        'KEY_F(3)': 'F3',
-        'KEY_F(4)': 'F4',
-        'KEY_F(5)': 'F5',
-        'KEY_F(6)': 'F6',
-        'KEY_F(7)': 'F7',
-        'KEY_F(8)': 'F8',
-        'KEY_F(9)': 'F9',
-        'KEY_F(10)': 'F10',
-        'KEY_F(11)': 'F11',
-        'KEY_F(12)': 'F12',
+        'BUTTON4_PRESSED': 'ScrollUp',
+        'BUTTON5_PRESSED': 'ScrollDown',
+        'REPORT_MOUSE_POSITION': 'ScrollDown',
+        '2097152': 'ScrollDown',
     }
 
+for i in range(1, 13):
+    d = visidata.vd.prettykeys_trdict
+    d[f'KEY_F({i})'] = f'F{i}'
+    d[f'KEY_F({i+12})'] = f'Shift+F{i}'
+    d[f'KEY_F({i+24})'] = f'Ctrl+F{i}'
+    d[f'KEY_F({i+36})'] = f'Ctrl+Shift+F{i}'
+    d[f'KEY_F({i+48})'] = f'Alt+F{i}'
+    d[f'KEY_F({i+60})'] = f'Alt+Shift+F{i}'
+
 
 @visidata.VisiData.api
 def prettykeys(vd, key):

visidata/loaders/__init__.py

@@ -0,0 +1,9 @@
+from visidata import vd, IndexSheet
+
+
+vd.option('load_lazy', False, 'load subsheets always (False) or lazily (True)')
+vd.option('skip', 0, 'skip N rows before header', replay=True)
+vd.option('header', 1, 'parse first N rows as column names', replay=True)
+
+IndexSheet.options.header = 0
+IndexSheet.options.skip = 0

visidata/loaders/_pandas.py

@@ -4,23 +4,57 @@ from visidata import VisiData, vd, Sheet, date, anytype, Path, options, Column,
 
 @VisiData.api
 def open_pandas(vd, p):
-    return PandasSheet(p.name, source=p)
+    return PandasSheet(p.base_stem, source=p)
 
 @VisiData.api
 def open_dta(vd, p):
-    return PandasSheet(p.name, source=p, filetype='stata')
+    return PandasSheet(p.base_stem, source=p, filetype='stata')
 
 VisiData.open_stata = VisiData.open_pandas
 
 for ft in 'feather gbq orc pickle sas stata'.split():
     funcname ='open_'+ft
     if not getattr(VisiData, funcname, None):
-        setattr(VisiData, funcname, lambda vd,p,ft=ft: PandasSheet(p.name, source=p, filetype=ft))
+        setattr(VisiData, funcname, lambda vd,p,ft=ft: PandasSheet(p.base_stem, source=p, filetype=ft))
 
+@VisiData.api
+@asyncthread
+def save_dta(vd, p, *sheets):
+    import pandas as pd
+    import numpy as np
+
+    # STATA is a one-sheet software
+    # Save only the first sheet
+    vs = sheets[0]
+
+    columns = [col.name for col in vs.visibleCols]
+    
+    # Get data types
+    types = list()
+    dispvals = next(vs.iterdispvals(format=True))
+    for col,_ in dispvals.items():
+        if col.type in [bool, int, float]:
+            types.append(col.type)
+        elif vd.isNumeric(col):
+            types.append(float)
+        else:
+            types.append(str)
+
+    # Populate numpy array
+    data = np.empty((vs.nRows, len(columns)), dtype=object)
+    for r_i, dispvals in enumerate(vs.iterdispvals(format=True)):
+        for c_i, v in enumerate(dispvals.values()):
+            data[r_i, c_i] = v
+
+    # Convert to pandas DataFrame and save
+    dtype = {col:t for col,t in zip(columns, types)}
+    df = pd.DataFrame(data, columns=columns)
+    df = df.astype(dtype)
+    df.to_stata(p, version=118, write_index=False)
 
 class DataFrameAdapter:
     def __init__(self, df):
-        import pandas as pd
+        pd = vd.importExternal('pandas')
         if not isinstance(df, pd.DataFrame):
             vd.fail('%s is not a dataframe' % type(df).__name__)
 
@@ -57,7 +91,7 @@ class PandasSheet(Sheet):
     '''
 
     def dtype_to_type(self, dtype):
-        import numpy as np
+        np = vd.importExternal('numpy')
         # Find the underlying numpy dtype for any pandas extension dtypes
         dtype = getattr(dtype, 'numpy_dtype', dtype)
         try:
@@ -74,7 +108,7 @@ class PandasSheet(Sheet):
 
     def read_tsv(self, path, **kwargs):
         'Partial function for reading TSV files using pd.read_csv'
-        import pandas as pd
+        pd = vd.importExternal('pandas')
         return pd.read_csv(path, sep='\t', **kwargs)
 
     @property
@@ -111,7 +145,7 @@ class PandasSheet(Sheet):
 
     @asyncthread
     def reload(self):
-        import pandas as pd
+        pd = vd.importExternal('pandas')
         if isinstance(self.source, pd.DataFrame):
             df = self.source
         elif isinstance(self.source, Path):
@@ -124,6 +158,12 @@ class PandasSheet(Sheet):
                 readfunc = getattr(pd, 'read_'+filetype) or vd.error('no pandas.read_'+filetype)
             # readfunc() handles binary and text open()
             df = readfunc(self.source, **options.getall('pandas_'+filetype+'_'))
+            # some read methods (html, for example) return a list of dataframes
+            if isinstance(df, list):
+                for idx, inner_df in enumerate(df[1:], start=1):
+                    vd.push(PandasSheet(f'{self.name}[{idx}]', source=inner_df))
+                df = df[0]
+                self.name += '[0]'
             if (filetype == 'pickle') and not isinstance(df, pd.DataFrame):
                 vd.fail('pandas loader can only unpickle dataframes')
         else:
@@ -169,7 +209,7 @@ class PandasSheet(Sheet):
         self.rows.sort_values(by=by_cols, ascending=ascending, inplace=True)
 
     def _checkSelectedIndex(self):
-        import pandas as pd
+        pd = vd.importExternal('pandas')
         if self._selectedMask.index is not self.df.index:
             # DataFrame was modified inplace, so the selection is no longer valid
             vd.status('pd.DataFrame.index updated, clearing {} selected rows'
@@ -224,7 +264,7 @@ class PandasSheet(Sheet):
             self.unselectRow(row)
 
     def clearSelected(self):
-        import pandas as pd
+        pd = vd.importExternal('pandas')
         self._selectedMask = pd.Series(False, index=self.df.index)
 
     def selectByIndex(self, start=None, end=None):
@@ -251,7 +291,7 @@ class PandasSheet(Sheet):
         matching rows to the selection. If unselect is True, remove from the
         active selection instead.
         '''
-        import pandas as pd
+        pd = vd.importExternal('pandas')
         case_sensitive = 'I' not in vd.options.regex_flags
         masks = pd.DataFrame([
             self.df[col.expr].astype(str).str.contains(pat=regex, case=case_sensitive, regex=True)
@@ -278,13 +318,13 @@ class PandasSheet(Sheet):
         DataFrame's dtypes.
         '''
 
-        import pandas as pd
+        pd = vd.importExternal('pandas')
         return pd.DataFrame({
             col: [None] * n for col in self.df.columns
         }).astype(self.df.dtypes.to_dict(), errors='ignore')
 
     def addRows(self, rows, index=None, undo=True):
-        import pandas as pd
+        pd = vd.importExternal('pandas')
         if index is None:
             self.df = self.df.append(pd.DataFrame(rows))
         else:
@@ -296,7 +336,7 @@ class PandasSheet(Sheet):
             vd.addUndo(self._deleteRows, range(index, index + len(rows)))
 
     def _deleteRows(self, which):
-        import pandas as pd
+        pd = vd.importExternal('pandas')
         self.df.drop(which, inplace=True)
         self.df.index = pd.RangeIndex(self.nRows)
         self._checkSelectedIndex()
@@ -306,7 +346,7 @@ class PandasSheet(Sheet):
         vd.addUndo(self._deleteRows, index or self.nRows - 1)
 
     def delete_row(self, rowidx):
-        import pandas as pd
+        pd = vd.importExternal('pandas')
         oldrow = self.df.iloc[rowidx:rowidx+1]
 
         # Use to_dict() here to work around an edge case when applying undos.
@@ -321,7 +361,7 @@ class PandasSheet(Sheet):
 
     def deleteBy(self, by):
         '''Delete rows for which func(row) is true.  Returns number of deleted rows.'''
-        import pandas as pd
+        pd = vd.importExternal('pandas')
         nRows = self.nRows
         vd.addUndo(setattr, self, 'df', self.df.copy())
         self.df = self.df[~by]
@@ -353,17 +393,17 @@ PandasSheet.addCommand(None, 'unselect-before', 'unselectByIndex(end=cursorRowIn
 PandasSheet.addCommand(None, 'stoggle-after', 'toggleByIndex(start=cursorRowIndex)', 'toggle selection of rows from cursor to bottom')
 PandasSheet.addCommand(None, 'select-after', 'selectByIndex(start=cursorRowIndex)', 'select all rows from cursor to bottom')
 PandasSheet.addCommand(None, 'unselect-after', 'unselectByIndex(start=cursorRowIndex)', 'unselect all rows from cursor to bottom')
-PandasSheet.addCommand(None, 'random-rows', 'nrows=int(input("random number to select: ", value=nRows)); vs=copy(sheet); vs.name=name+"_sample"; vs.rows=DataFrameAdapter(sheet.df.sample(nrows or nRows)); vd.push(vs)', 'open duplicate sheet with a random population subset of N rows'),
+PandasSheet.addCommand(None, 'random-rows', 'nrows=int(input("random number to select: ", value=nRows)); vs=copy(sheet); vs.name=name+"_sample"; vs.rows=DataFrameAdapter(sheet.df.sample(nrows or nRows)); vd.push(vs)', 'open duplicate sheet with a random population subset of N rows')
 
 # Handle the regex selection family of commands through a single method,
 # since the core logic is shared
-PandasSheet.addCommand('|', 'select-col-regex', 'selectByRegex(regex=input("select regex: ", type="regex", defaultLast=True), columns=[cursorCol])', 'select rows matching regex in current column')
-PandasSheet.addCommand('\\', 'unselect-col-regex', 'selectByRegex(regex=input("select regex: ", type="regex", defaultLast=True), columns=[cursorCol], unselect=True)', 'unselect rows matching regex in current column')
-PandasSheet.addCommand('g|', 'select-cols-regex', 'selectByRegex(regex=input("select regex: ", type="regex", defaultLast=True), columns=visibleCols)', 'select rows matching regex in any visible column')
-PandasSheet.addCommand('g\\', 'unselect-cols-regex', 'selectByRegex(regex=input("select regex: ", type="regex", defaultLast=True), columns=visibleCols, unselect=True)', 'unselect rows matching regex in any visible column')
+PandasSheet.addCommand('|', 'select-col-regex', 'selectByRegex(regex=inputRegex("select regex: ", defaultLast=True), columns=[cursorCol])', 'select rows matching regex in current column')
+PandasSheet.addCommand('\\', 'unselect-col-regex', 'selectByRegex(regex=inputRegex("select regex: ", defaultLast=True), columns=[cursorCol], unselect=True)', 'unselect rows matching regex in current column')
+PandasSheet.addCommand('g|', 'select-cols-regex', 'selectByRegex(regex=inputRegex("select regex: ", defaultLast=True), columns=visibleCols)', 'select rows matching regex in any visible column')
+PandasSheet.addCommand('g\\', 'unselect-cols-regex', 'selectByRegex(regex=inputRegex("select regex: ", defaultLast=True), columns=visibleCols, unselect=True)', 'unselect rows matching regex in any visible column')
 
 # Override with a pandas/dataframe-aware implementation
-PandasSheet.addCommand('"', 'dup-selected', 'vs=PandasSheet(sheet.name, "selectedref", source=selectedRows.df); vd.push(vs)', 'open duplicate sheet with only selected rows'),
+PandasSheet.addCommand('"', 'dup-selected', 'vs=PandasSheet(sheet.name, "selectedref", source=selectedRows.df); vd.push(vs)', 'open duplicate sheet with only selected rows')
 
 vd.addGlobals({
     'PandasSheet': PandasSheet,

visidata/loaders/api_airtable.py

@@ -0,0 +1,70 @@
+import re
+import os
+
+from visidata import vd, date, asyncthread, VisiData, Progress, Sheet, Column, ItemColumn, deduceType, TypedWrapper, setitem
+
+
+vd.option('airtable_auth_token', '', 'Airtable API key from https://airtable.com/account')
+
+airtable_regex = r'^https://airtable.com/(app[A-Za-z0-9]+)/(tbl[A-Za-z0-9]+)/?(viw[A-z0-9]+)?'
+
+@VisiData.api
+def guessurl_airtable(vd, p, response):
+    m = re.search(airtable_regex, p.given)
+    if m:
+        return dict(filetype='airtable', _likelihood=10)
+
+
+@VisiData.api
+def open_airtable(vd, p):
+    pyairtable = vd.importExternal('pyairtable')
+
+    token = os.environ.get('AIRTABLE_AUTH_TOKEN') or vd.options.airtable_auth_token
+    if not token:
+        vd.requireOptions('airtable_auth_token', help='https://support.airtable.com/docs/creating-and-using-api-keys-and-access-tokens')
+
+    m = re.search(airtable_regex, p.given)
+    if not m:
+        vd.fail('invalid airtable url')
+
+    app, tbl, viw = m.groups()
+    return AirtableSheet('airtable', source=p,
+                         airtable_auth_token=token,
+                         airtable_base=app,
+                         airtable_table=tbl,
+                         airtable_view=viw)
+
+
+class AirtableSheet(Sheet):
+    guide = '''
+        # Airtable
+        This sheet is a read-only download of all records in a table at _airtable.com_.
+    '''
+    rowtype = 'records'  # rowdef: dict
+
+    columns = [
+        ItemColumn('id', 'id', type=str, width=0),
+        ItemColumn('createdTime', 'createdTime', type=date, width=0)
+    ]
+
+    def iterload(self):
+        self.fields = set()
+
+        for page in self.api.iterate(self.airtable_base, self.airtable_table, view=self.airtable_view):
+            for row in page:
+                yield row
+
+                for field, value in row['fields'].items():
+                    if field not in self.fields:
+                        col = ItemColumn('fields.'+field, type=deduceType(value))
+                        self.addColumn(col)
+                        self.fields.add(field)
+
+    def newRow(self):
+        return AttrDict(fields=AttrDict())
+
+
+@AirtableSheet.lazy_property
+def api(self):
+    import pyairtable
+    return pyairtable.Api(self.airtable_auth_token)

visidata/loaders/api_bitio.py

@@ -0,0 +1,102 @@
+
+from visidata import *
+
+vd.option('bitio_api_key', '', 'API key')
+
+@VisiData.api
+def new_bitio(vd, p):
+    vd.importExternal('bitdotio')
+    vd.requireOptions('bitio_api_key', help='https://docs.bit.io/docs/connecting-via-the-api')
+    return BitioReposSheet(p.name, source=p)
+
+vd.openhttp_bitio = vd.new_bitio
+
+@VisiData.lazy_property
+def bitio_client(vd):
+    import bitdotio
+    return bitdotio.bitdotio(vd.options.bitio_api_key)
+
+@VisiData.api
+def bitio_api(vd, path, method, **kwargs):
+    t = vd.bitio_client.api_client.call_api(path, method,
+            header_params={'Accept': 'application/json',
+                           'Authorization': 'Bearer '+vd.bitio_client.access_token,
+                           'Content-Type': 'application/json'}, async_req=True, body=kwargs)
+    if not t.successful():
+        vd.warning(resp['Reason'])
+    return t.get()
+
+
+class BitioReposSheet(Sheet):
+    rowtype = 'repos'
+    columns = [
+        AttrColumn('name'),
+        AttrColumn('creator'),
+        AttrColumn('owner'),
+        AttrColumn('bytes', type=int, width=0),
+        AttrColumn('collaborators'),
+        AttrColumn('description'),
+        AttrColumn('documentation'),
+        AttrColumn('endpoints'),
+        AttrColumn('is_private'),
+        AttrColumn('license'),
+        AttrColumn('query_count'),
+        AttrColumn('stars'),
+        AttrColumn('tables'),
+        AttrColumn('url'),
+        AttrColumn('watchers'),
+    ]
+    defer = True
+    def iterload(self):
+        yield from vd.bitio_client.list_repos(self.source.name)
+
+
+    @asyncthread
+    def putChanges(self):
+        adds, mods, dels = self.getDeferredChanges()
+        for row in dels.values():
+            vd.bitio_api(f'/users/{self.source.name}/repos/{row.name}/', 'DELETE')
+
+        for row, rowmods in Progress(list(mods.values()), gerund="updating"):
+            kwargs = {col.name: val for col, val in rowmods.items()}
+            vd.bitio_api(f'/users/{self.source.name}/repos/{row.name}/', 'PATCH', **kwargs)
+
+    def openRow(self, row):
+        return BitioRepoSheet(self.source.name, row.name, source=row)
+
+
+class BitioRepoSheet(Sheet):
+    rowtype = 'tables'
+    columns = [
+        AttrColumn('url'),
+        AttrColumn('current_name'),
+        AttrColumn('description'),
+        AttrColumn('columns'),
+        AttrColumn('num_records', type=int),
+        AttrColumn('bytes', type=int),
+        AttrColumn('repo'),
+        AttrColumn('documentation'),
+    ]
+    def iterload(self):
+        yield from vd.bitio_client.list_tables(self.source.owner.split('/')[-2], self.source.name)
+
+    def openRow(self, row):
+        username = row.repo.split('/')[-4]
+        repo = row.repo.split('/')[-2]
+        return BitioTable(username, repo, source=row)
+
+
+class BitioTable(Sheet):
+    def iterload(self):
+        username = self.source.repo.split('/')[-4]
+        repo = self.source.repo.split('/')[-2]
+        conn = vd.bitio_client.get_connection()
+        with conn.cursor() as cur:
+            cur.execute(f'SELECT * FROM "{username}/{repo}"."{self.source.current_name}"')
+            r = cur.fetchone()
+            if r:
+                yield r
+            self.columns = []
+            for c in vd.postgresGetColumns(cur):
+                self.addColumn(c)
+            yield from cur