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

visidata/loaders/ttf.py

@@ -3,7 +3,7 @@ from visidata import VisiData, vd, Sheet, Column, Progress, ColumnAttr, ColumnIt
 
 @VisiData.api
 def open_ttf(vd, p):
-    return TTFTablesSheet(p.name, source=p)
+    return TTFTablesSheet(p.base_stem, source=p)
 
 vd.open_otf = vd.open_ttf
 
@@ -24,9 +24,9 @@ class TTFTablesSheet(Sheet):
         return TTFGlyphsSheet(self.name+'_glyphs', source=self, sourceRows=[row], ttf=self.ttf)
 
     def iterload(self):
-        import fontTools.ttLib
+        fontTools = vd.importExternal('fontTools.ttLib', 'fonttools')
 
-        self.ttf = fontTools.ttLib.TTFont(str(self.source), 0, allowVID=0, ignoreDecompileErrors=True, fontNumber=-1)
+        self.ttf = fontTools.TTFont(str(self.source), 0, allowVID=0, ignoreDecompileErrors=True, fontNumber=-1)
         for cmap in self.ttf["cmap"].tables:
             yield cmap
 
@@ -53,20 +53,21 @@ class TTFGlyphsSheet(Sheet):
 
 
 def makePen(*args, **kwargs):
-    try:
-        from fontTools.pens.basePen import BasePen
-    except ImportError as e:
-        vd.error('fonttools not installed')
+    fontTools = vd.importExternal('fontTools', 'fonttools')
+    from fontTools.pens.basePen import BasePen
 
     class GlyphPen(InvertedCanvas, BasePen):
         aspectRatio = 1.0
         def __init__(self, name, **kwargs):
             super().__init__(name, **kwargs)
+            self.path_firstxy = None
             self.lastxy = None
             self.attr = self.plotColor(('glyph',))
 
         def _moveTo(self, xy):
             self.lastxy = xy
+            if self.path_firstxy is None:
+                self.path_firstxy = xy
 
         def _lineTo(self, xy):
             x1, y1 = self.lastxy
@@ -74,6 +75,17 @@ def makePen(*args, **kwargs):
             self.line(x1, y1, x2, y2, self.attr)
             self._moveTo(xy)
 
+        def _closePath(self):
+            if self.path_firstxy:
+                if (self.path_firstxy != self.lastxy):
+                    self._lineTo(self.path_firstxy)
+                self.path_firstxy = None
+            self.lastxy = None
+
+        def _endPath(self):
+            self.path_firstxy = None
+            self.lastxy = None
+
         def _curveToOne(self, xy1, xy2, xy3):
             vd.error('NotImplemented')
 

visidata/loaders/unzip_http.py

@@ -26,6 +26,7 @@ import struct
 import fnmatch
 import pathlib
 import urllib.parse
+from visidata import vd
 
 
 __version__ = '0.5.1'
@@ -98,7 +99,7 @@ class RemoteZipFile:
     magic_eocd = b'\x50\x4b\x05\x06'
 
     def __init__(self, url):
-        import urllib3
+        urllib3 = vd.importExternal('urllib3')
         self.url = url
         self.http = urllib3.PoolManager()
         self.zip_size = 0
@@ -185,7 +186,7 @@ class RemoteZipFile:
 
             outpath = path/member
             os.makedirs(outpath.parent, exist_ok=True)
-            with self.open(member) as fpin:
+            with self._open(member) as fpin:
                 with open(path/member, mode='wb') as fpout:
                     while True:
                         r = fpin.read(65536)
@@ -205,7 +206,7 @@ class RemoteZipFile:
             if any(fnmatch.fnmatch(f.filename, g) for g in globs):
                 yield f
 
-    def open(self, fn):
+    def _open(self, fn):
         if isinstance(fn, str):
             f = list(self.matching_files(fn))
             if not f:
@@ -226,8 +227,8 @@ class RemoteZipFile:
         else:
             error(f'unknown compression method {method}')
 
-    def open_text(self, fn):
-        return io.TextIOWrapper(self.open(fn))
+    def open(self, fn):
+        return io.TextIOWrapper(self._open(fn))
 
 
 class RemoteZipStream(io.RawIOBase):

visidata/loaders/usv.py

@@ -4,7 +4,7 @@ from visidata import Sheet, TsvSheet, options, vd, VisiData
 
 @VisiData.api
 def open_usv(vd, p):
-    return TsvSheet(p.name, source=p, delimiter='\u241f', row_delimiter='\u241e')
+    return TsvSheet(p.base_stem, source=p, delimiter='\u241f', row_delimiter='\u241e')
 
 @VisiData.api
 def save_usv(vd, p, vs):

visidata/loaders/vcf.py

@@ -1,11 +1,11 @@
-from visidata import VisiData, Column, getitemdef, PythonSheet, asyncthread
+from visidata import VisiData, Column, getitemdef, PythonSheet, asyncthread, vd
 
 
 # requires (deb): libbz2-dev libcurl4-openssl-dev liblzma-dev
 
 @VisiData.api
 def open_vcf(vd, p):
-    return VcfSheet(p.name, source=p)
+    return VcfSheet(p.base_stem, source=p)
 
 def unbox(col, row):
     v = getitemdef(row, col.expr)
@@ -20,22 +20,22 @@ class VcfSheet(PythonSheet):
     rowtype = 'cards'
     @asyncthread
     def reload(self):
-        import vobject
+        vobject = vd.importExternal('vobject')
         self.rows = []
         self.columns = []
 
         addedCols = set()
         lines = []
-        for line in self.source.open_text(encoding=self.options.encoding):
-            lines.append(line)
-            if line.startswith('END:'):
-                row = vobject.readOne('\n'.join(lines))
-                for k, v in row.contents.items():
-                    if v and str(v[0].value).startswith('(None)'):
-                        continue
-                    if not k in addedCols:
-                        addedCols.add(k)
-                        self.addColumn(Column(k, expr=k, getter=unbox))
-                self.addRow(row.contents)
-                lines = []
-
+        with self.open_text_source() as fp:
+            for line in fp:
+                lines.append(line)
+                if line.startswith('END:'):
+                    row = vobject.readOne('\n'.join(lines))
+                    for k, v in row.contents.items():
+                        if v and str(v[0].value).startswith('(None)'):
+                            continue
+                        if not k in addedCols:
+                            addedCols.add(k)
+                            self.addColumn(Column(k, expr=k, getter=unbox))
+                    self.addRow(row.contents)
+                    lines = []

visidata/loaders/vds.py

@@ -2,37 +2,40 @@
 
 import json
 
-from visidata import VisiData, JsonSheet, Progress, IndexSheet, SettableColumn, ItemColumn
+from visidata import VisiData, JsonSheet, Progress, IndexSheet, SettableColumn, ItemColumn, ExprColumn
 
 
 NL='\n'
 
 @VisiData.api
 def open_vds(vd, p):
-    return VdsIndexSheet(p.name, source=p)
+    return VdsIndexSheet(p.base_stem, source=p)
 
 
 @VisiData.api
 def save_vds(vd, p, *sheets):
     'Save in custom VisiData format, preserving columns and their attributes.'
 
-    with p.open_text(mode='w', encoding='utf-8') as fp:
+    with p.open(mode='w', encoding='utf-8') as fp:
         for vs in sheets:
             # class and attrs for vs
             d = { 'name': vs.name, }
             fp.write('#'+json.dumps(d)+NL)
 
             # class and attrs for each column in vs
-            for col in vs.visibleCols:
+            for col in vs.columns:
                 d = col.__getstate__()
                 if isinstance(col, SettableColumn):
                     d['col'] = 'Column'
+                elif isinstance(col, ItemColumn):
+                    d['col'] = 'Column'
+                    d['expr'] = col.name  #2037  override expr
                 else:
                     d['col'] = type(col).__name__
                 fp.write('#'+json.dumps(d)+NL)
 
             with Progress(gerund='saving'):
-                for row in vs.iterdispvals(*vs.visibleCols, format=False):
+                for row in vs.iterdispvals(*vs.columns, format=False):
                     d = {col.name:val for col, val in row.items()}
                     fp.write(json.dumps(d, default=str)+NL)
 
@@ -40,7 +43,7 @@ def save_vds(vd, p, *sheets):
 class VdsIndexSheet(IndexSheet):
     def iterload(self):
         vs = None
-        with self.source.open_text(encoding='utf-8') as fp:
+        with self.source.open(encoding='utf-8') as fp:
             line = fp.readline()
             while line:
                 if line.startswith('#{'):
@@ -59,7 +62,7 @@ class VdsSheet(JsonSheet):
         self.colnames = {}
         self.columns = []
 
-        with self.source.open_text(encoding='utf-8') as fp:
+        with self.source.open(encoding='utf-8') as fp:
             fp.seek(self.source_fpos)
 
             # consume all metadata, create columns

visidata/loaders/vdx.py

@@ -1,10 +1,12 @@
+import re
+
 import visidata
-from visidata import VisiData, CommandLogBase, BaseSheet, Sheet, AttrDict
+from visidata import VisiData, CommandLogBase, BaseSheet, Sheet, AttrDict, Progress
 
 
 @VisiData.api
 def open_vdx(vd, p):
-    return CommandLogSimple(p.name, source=p, precious=True)
+    return CommandLogSimple(p.base_stem, source=p, precious=True)
 
 
 class CommandLogSimple(CommandLogBase, Sheet):
@@ -20,16 +22,16 @@ class CommandLogSimple(CommandLogBase, Sheet):
 
 @VisiData.api
 def save_vdx(vd, p, *vsheets):
-    with p.open_text(mode='w', encoding=vsheets[0].options.encoding) as fp:
+    with p.open(mode='w', encoding=vsheets[0].options.save_encoding) as fp:
         fp.write(f"# {visidata.__version_info__}\n")
         for vs in vsheets:
             prevrow = None
             for r in vs.rows:
                 if prevrow is not None and r.sheet and prevrow.sheet != r.sheet:
                     fp.write(f'sheet {r.sheet}\n')
-                if r.col and (prevrow.col != r.col or prevrow is None):
+                if r.col and (prevrow is None or prevrow.col != r.col):
                     fp.write(f'col {r.col}\n')
-                if r.row and (prevrow.row != r.row or prevrow is None):
+                if r.row and (prevrow is None or prevrow.row != r.row):
                     fp.write(f'row {r.row}\n')
 
                 line = r.longname
@@ -40,6 +42,29 @@ def save_vdx(vd, p, *vsheets):
                 prevrow = r
 
 
+@VisiData.api
+def runvdx(vd, vdx:str):
+    for line in Progress(vdx.splitlines()):
+        vs = vd.sheet or Sheet()
+        vs.ensureLoaded()
+        line = line.strip()
+        if not line or line[0] == '#':
+            continue
+
+        m = re.match(r'^(\+(\S+) )?(\S+)(.*)$', line)
+        if not m:
+            print('bad:', line)
+            continue
+
+        _, pos, longname, rest = m.groups()
+        vd.currentReplayRow = AttrDict(longname=longname, input=rest)
+        if pos:
+            vd.moveToPos(vd.sheets, *vd.parsePos(pos))
+        print(vs.name, longname)
+        vs.execCommand(longname)
+        vd.sync()
+
+
 BaseSheet.addCommand('', 'sheet', 'n=input("sheet to jump to: "); vd.push(vd.getSheet(n) or fail(f"no such sheet {n}"))', 'jump to named sheet')
 BaseSheet.addCommand('', 'col', 'n=input("column to go to: "); moveToCol(n) or fail(f"no such column {n}")', 'move to named/numbered col')
 BaseSheet.addCommand('', 'row', 'n=input("row to go to: "); moveToRow(n) or fail(f"no such row {n}")', 'move to named/numbered row')

visidata/loaders/xlsb.py

@@ -2,13 +2,20 @@ from visidata import vd, IndexSheet, VisiData
 
 'Requires visidata/deps/pyxlsb fork'
 
+@VisiData.api
+def guess_xls(vd, p):
+    if p.open_bytes().read(16).startswith(b'\xD0\xCF\x11\xE0\xA1\xB1\x1A\xE1'):
+        return dict(filetype='xlsb', _likelihood=10)
+
 
 @VisiData.api
 def open_xlsb(vd, p):
-    return XlsbIndex(p.name, source=p)
+    return XlsbIndex(p.base_stem, source=p)
+
 
 class XlsbIndex(IndexSheet):
     def iterload(self):
+        vd.importExternal('pyxlsb', '-e git+https://github.com/saulpw/pyxlsb.git@visidata#egg=pyxlsb')
         from pyxlsb import open_workbook
 
         wb = open_workbook(str(self.source))

visidata/loaders/xlsx.py

@@ -1,8 +1,11 @@
 import itertools
 import copy
 import datetime
+import re
+from colorsys import rgb_to_hls, hls_to_rgb
 
 from visidata import VisiData, vd, Sheet, Column, Progress, IndexSheet, ColumnAttr, SequenceSheet, AttrDict, AttrColumn
+from visidata import CellColorizer, getattrdeep, rgb_to_attr
 from visidata.type_date import date
 
 
@@ -10,11 +13,11 @@ vd.option('xlsx_meta_columns', False, 'include columns for cell objects, font co
 
 @VisiData.api
 def open_xls(vd, p):
-    return XlsIndexSheet(p.name, source=p)
+    return XlsIndexSheet(p.base_stem, source=p)
 
 @VisiData.api
 def open_xlsx(vd, p):
-    return XlsxIndexSheet(p.name, source=p)
+    return XlsxIndexSheet(p.base_stem, source=p)
 
 class XlsxIndexSheet(IndexSheet):
     'Load XLSX file (in Excel Open XML format).'
@@ -29,16 +32,20 @@ class XlsxIndexSheet(IndexSheet):
     nKeys = 1
 
     def iterload(self):
-        import openpyxl
+        openpyxl = vd.importExternal('openpyxl')
         self.workbook = openpyxl.load_workbook(str(self.source), data_only=True, read_only=True)
         for sheetname in self.workbook.sheetnames:
             src = self.workbook[sheetname]
-            yield XlsxSheet(self.name, sheetname, source=src)
+            yield XlsxSheet(self.name, sheetname, source=src, workbook=self.workbook)
 
 
 class XlsxSheet(SequenceSheet):
     # rowdef: AttrDict of column_letter to cell
+    colorizers = [
+        CellColorizer(5, None, lambda s,c,r,v: c and r and s.colorize_xlsx_cell(c,r))
+    ]
     def setCols(self, headerrows):
+        vd.importExternal('openpyxl')
         from openpyxl.utils.cell import get_column_letter
         self.columns = []
         self._rowtype = AttrDict
@@ -55,7 +62,7 @@ class XlsxSheet(SequenceSheet):
         for i, colnamelines in enumerate(itertools.zip_longest(*headers, fillvalue='')):
             colnamelines = ['' if c is None else c for c in colnamelines]
             column_name = ''.join(map(str, colnamelines))
-            self.addColumn(AttrColumn(column_name, column_letters[i] + '.value'))
+            self.addColumn(AttrColumn(column_name, column_letters[i] + '.value', column_letter=column_letters[i]))
             self.addXlsxMetaColumns(column_letters[i], column_name)
 
     def addRow(self, row, index=None):
@@ -65,6 +72,7 @@ class XlsxSheet(SequenceSheet):
             self.addXlsxMetaColumns(column_letter, column_letter)
 
     def iterload(self):
+        vd.importExternal('openpyxl')
         from openpyxl.utils.cell import get_column_letter
         worksheet = self.source
         for row in Progress(worksheet.iter_rows(), total=worksheet.max_row or 0):
@@ -73,13 +81,16 @@ class XlsxSheet(SequenceSheet):
     def addXlsxMetaColumns(self, column_letter, column_name):
         if self.options.xlsx_meta_columns:
             self.addColumn(
-                    AttrColumn(column_name + '_cellPyObj', column_letter))
+                    AttrColumn(column_name + '_cellPyObj', column_letter, column_letter=column_letter))
             self.addColumn(
                     AttrColumn(column_name + '_fontColor',
-                        column_letter + '.font.color.value'))
+                        column_letter + '.font.color.value', column_letter=column_letter))
             self.addColumn(
                     AttrColumn(column_name + '_fillColor', column_letter +
-                        '.fill.start_color.value'))
+                        '.fill.start_color.value', column_letter=column_letter))
+            self.addColumn(Column(column_name + '_colorizer', width=0,
+                           column_letter=column_letter,
+                           getter=lambda c,r: c.sheet.colorize_xlsx_cell(c,r)))
 
     def paste_after(self, rowidx):
         to_paste = list(copy.copy(r) for r in reversed(vd.memory.cliprows))
@@ -97,7 +108,7 @@ class XlsIndexSheet(IndexSheet):
     ]
     nKeys = 1
     def iterload(self):
-        import xlrd
+        xlrd = vd.importExternal('xlrd')
         self.workbook = xlrd.open_workbook(str(self.source))
         for sheetname in self.workbook.sheet_names():
             yield XlsSheet(self.name, sheetname, source=self.workbook.sheet_by_name(sheetname))
@@ -123,11 +134,26 @@ def xls_name(vs):
 
 @VisiData.api
 def save_xlsx(vd, p, *sheets):
-    import openpyxl
+    openpyxl = vd.importExternal('openpyxl')
 
     wb = openpyxl.Workbook()
     wb.remove_sheet(wb['Sheet'])
 
+    def _convert_save_row(dispvals:dict, replace_illegal=False) -> list:
+        row = []
+        for col, v in dispvals.items():
+            if v is None:
+                v = ""
+            elif col.type == date:
+                v = datetime.datetime.fromtimestamp(int(v.timestamp()))
+            elif not vd.isNumeric(col):
+                v = str(v)
+                if replace_illegal:
+                    v = re.sub(openpyxl.cell.cell.ILLEGAL_CHARACTERS_RE, ' ', v)
+
+            row.append(v)
+        return row
+
     for vs in sheets:
         if vs.xls_name != vs.names[-1]:
             vd.warning(f'saving {vs.name} as {vs.xls_name}')
@@ -137,28 +163,21 @@ def save_xlsx(vd, p, *sheets):
         ws.append(headers)
 
         for dispvals in vs.iterdispvals(format=False):
-
-            row = []
-            for col, v in dispvals.items():
-                if v is None:
-                    v = ""
-                elif col.type == date:
-                    v = datetime.datetime.fromtimestamp(int(v.timestamp()))
-                elif not vd.isNumeric(col):
-                    v = str(v)
-                row.append(v)
-
-            ws.append(row)
+            row = _convert_save_row(dispvals)
+            try:
+                ws.append(row)
+            except openpyxl.utils.exceptions.IllegalCharacterError as e:
+                row = _convert_save_row(dispvals, replace_illegal=True)  #1402
+                ws.append(row)
 
     wb.active = ws
 
     wb.save(filename=p)
-    vd.status(f'{p} save finished')
 
 
 @VisiData.api
 def save_xls(vd, p, *sheets):
-    import xlwt
+    xlwt = vd.importExternal('xlwt')
 
     wb = xlwt.Workbook()
 
@@ -175,4 +194,105 @@ def save_xls(vd, p, *sheets):
                 ws1.write(r_i, c_i, v)
 
     wb.save(p)
-    vd.status(f'{p} save finished')
+
+
+# from https://stackoverflow.com/a/65426130
+
+RGBMAX = 255
+HLSMAX = 240
+
+@XlsxSheet.api
+def colorize_xlsx_cell(sheet, col, row):
+    fg = getattrdeep(row, col.column_letter+'.font.color', None)
+    bg = getattrdeep(row, col.column_letter+'.fill.start_color', None)
+    fg = sheet.xlsx_color_to_xterm256(fg)
+    bg = sheet.xlsx_color_to_xterm256(bg)
+
+    if bg == '-1' or fg == '-1':
+        fg, bg = '-1', '-1'
+
+    return f'{fg} on {bg}'
+
+@XlsxSheet.api
+def xlsx_color_to_xterm256(sheet, color) -> str:
+    if not color:
+        return ''
+
+    if color.type == 'rgb':
+        s = color.value
+        if isinstance(s, int):
+            return str(s)
+
+        a,r,g,b = s[0:2], s[2:4], s[4:6], s[6:8]
+        return rgb_to_attr(int(r, 16), int(g, 16), int(b, 16), int(a, 16))
+
+    if color.type == 'theme':
+        return sheet.theme_and_tint_to_rgb(color.value, color.tint)
+    else:
+        return str(color.value)
+
+@XlsxSheet.api
+def theme_and_tint_to_rgb(sheet, theme, tint) -> str:
+    """Given a workbook, a theme number and a tint return a xterm256 color number"""
+    rgb = sheet.theme_colors[theme]
+    h, l, s = rgb_to_ms_hls(rgb)
+    r, g, b = ms_hls_to_rgb(h, tint_luminance(tint, l), s)
+
+    return rgb_to_attr(r*256, g*256, b*256)
+
+@XlsxSheet.lazy_property
+def theme_colors(sheet):
+    """Gets theme colors from the workbook"""
+    # see: https://groups.google.com/forum/#!topic/openpyxl-users/I0k3TfqNLrc
+    from openpyxl.xml.functions import QName, fromstring
+    xlmns = 'http://schemas.openxmlformats.org/drawingml/2006/main'
+    root = fromstring(sheet.workbook.loaded_theme)
+    themeEl = root.find(QName(xlmns, 'themeElements').text)
+    colorSchemes = themeEl.findall(QName(xlmns, 'clrScheme').text)
+    firstColorScheme = colorSchemes[0]
+
+    theme_colors = []
+
+    for c in ['lt1', 'dk1', 'lt2', 'dk2', 'accent1', 'accent2', 'accent3', 'accent4', 'accent5', 'accent6']:
+        accent = firstColorScheme.find(QName(xlmns, c).text)
+        for i in list(accent): # walk all child nodes, rather than assuming [0]
+            if 'window' in i.attrib['val']:
+                theme_colors.append(i.attrib['lastClr'])
+            else:
+                theme_colors.append(i.attrib['val'])
+
+    return theme_colors
+
+def rgb_to_ms_hls(red, green=None, blue=None):
+    """Converts rgb values in range (0,1) or a hex string of the form '[#aa]rrggbb' to HLSMAX based HLS, (alpha values are ignored)"""
+    if green is None:
+        if isinstance(red, str):
+            if len(red) > 6:
+                red = red[-6:]  # Ignore preceding '#' and alpha values
+            blue = int(red[4:], 16) / RGBMAX
+            green = int(red[2:4], 16) / RGBMAX
+            red = int(red[0:2], 16) / RGBMAX
+        else:
+            red, green, blue = red
+    h, l, s = rgb_to_hls(red, green, blue)
+    return (int(round(h * HLSMAX)), int(round(l * HLSMAX)), int(round(s * HLSMAX)))
+
+def ms_hls_to_rgb(hue, lightness=None, saturation=None):
+    """Converts HLSMAX based HLS values to rgb values in the range (0,1)"""
+    if lightness is None:
+        hue, lightness, saturation = hue
+    return hls_to_rgb(hue / HLSMAX, lightness / HLSMAX, saturation / HLSMAX)
+
+def rgb_to_hex(red, green=None, blue=None):
+    """Converts (0,1) based RGB values to a hex string 'rrggbb'"""
+    if green is None:
+        red, green, blue = red
+    return ('%02x%02x%02x' % (int(round(red * RGBMAX)), int(round(green * RGBMAX)), int(round(blue * RGBMAX)))).upper()
+
+def tint_luminance(tint, lum):
+    """Tints a HLSMAX based luminance"""
+    # See: http://ciintelligence.blogspot.co.uk/2012/02/converting-excel-theme-color-and-tint.html
+    if tint < 0:
+        return int(round(lum * (1.0 + tint)))
+    else:
+        return int(round(lum * (1.0 - tint) + (HLSMAX - HLSMAX * (1.0 - tint))))

visidata/loaders/xml.py

@@ -1,11 +1,12 @@
-from visidata import VisiData, vd, Sheet, options, Column, Progress, setitem, ColumnAttr, vlen, RowColorizer, Path, copy
+from copy import copy
+from visidata import VisiData, vd, Sheet, options, Column, Progress, setitem, ColumnAttr, vlen, RowColorizer, Path
 
 vd.option('xml_parser_huge_tree', True, 'allow very deep trees and very long text content')
 
 
 @VisiData.api
 def open_xml(vd, p):
-    return XmlSheet(p.name, source=p)
+    return XmlSheet(p.base_stem, source=p)
 
 VisiData.open_svg = VisiData.open_xml
 
@@ -47,9 +48,11 @@ class XmlSheet(Sheet):
 
     def iterload(self):
         if isinstance(self.source, Path):
+            vd.importExternal('lxml')
             from lxml import etree, objectify
             p = etree.XMLParser(**self.options.getall('xml_parser_'))
-            self.root = etree.parse(self.source.open_text(encoding=self.options.encoding), parser=p)
+            with self.open_text_source() as fp:
+                self.root = etree.parse(fp, parser=p)
             objectify.deannotate(self.root, cleanup_namespaces=True)
         else: #        elif isinstance(self.source, XmlElement):
             self.root = self.source

visidata/loaders/xword.py

@@ -8,13 +8,13 @@ vd.option('color_xword_active', 'green', 'color of active clue')
 
 @VisiData.api
 def open_puz(vd, p):
-    return PuzSheet(p.name, source=p)
+    return PuzSheet(p.base_stem, source=p)
 
 @VisiData.api
 def open_xd(vd, p):
     if p.is_dir():
-        return CrosswordsSheet(p.name, source=p)
-    return CrosswordSheet(p.name, source=p)
+        return CrosswordsSheet(p.base_stem, source=p)
+    return CrosswordSheet(p.base_stem, source=p)
 
 
 @VisiData.api
@@ -96,7 +96,7 @@ class PuzSheet(CrosswordSheet):
 
 @VisiData.api
 def save_xd(vd, p, vs):
-    with p.open_text(mode='w', encoding='utf-8') as fp:
+    with p.open(mode='w', encoding='utf-8') as fp:
         fp.write(vs.xd.to_unicode())
 
 

visidata/loaders/yaml.py

@@ -1,19 +1,29 @@
 from itertools import chain
 
-from visidata import VisiData, Progress, JsonSheet
+from visidata import VisiData, Progress, JsonSheet, vd
 
 
 @VisiData.api
 def open_yml(vd, p):
-    return YamlSheet(p.name, source=p)
+    return YamlSheet(p.base_stem, source=p)
 
 VisiData.open_yaml = VisiData.open_yml
 
 class YamlSheet(JsonSheet):
     def iterload(self):
-        import yaml
-        with self.source.open_text() as fp:
-            documents = yaml.safe_load_all(fp)
+        yaml = vd.importExternal('yaml', 'PyYAML')
+
+        class PrettySafeLoader(yaml.SafeLoader):
+            def construct_python_tuple(self, node):
+                return tuple(self.construct_sequence(node))
+
+        PrettySafeLoader.add_constructor(
+            u'tag:yaml.org,2002:python/tuple',
+            PrettySafeLoader.construct_python_tuple
+        )
+
+        with self.source.open() as fp:
+            documents = yaml.load_all(fp, PrettySafeLoader)
 
             self.columns = []
             self._knownKeys.clear()