// ----------------------------------------------------------------------------
// Filter implementation
-// A cell represents a segment of text delineated by tabs, form-feed,
+// A cell represents a segment of text delineated by tabs, formfeed,
// or newline chars. The text itself is stored in a separate buffer;
// cell only describes the segment's size in bytes, its width in runes,
// and whether it's an htab ('\t') or vtab ('\v') terminated call.
// are simply passed through. The widths of tags and entities are
// assumed to be zero (tags) and one (entities) for formatting purposes.
//
-// The form feed character ('\f') acts like a newline but it also
+// The formfeed character ('\f') acts like a newline but it also
// terminates all columns in the current line (effectively calling
// Flush). Cells in the next line start new columns. Unless found
-// inside an HTML tag, form feed characters appear as newlines in
+// inside an HTML tag, formfeed characters appear as newlines in
// the output.
//
// The Writer must buffer input internally, because proper spacing
flags uint;
// current state
- buf bytes.Buffer; // collected text w/o tabs, newlines, or form feed chars
+ buf bytes.Buffer; // collected text w/o tabs, newlines, or formfeed chars
pos int; // buffer position up to which width of incomplete cell has been computed
cell cell; // current incomplete cell; cell.width is up to buf[pos] w/o ignored sections
html_char byte; // terminating char of html tag/entity, or 0 ('>', ';', or 0)
// Internal representation (current state):
//
-// - all text written is appended to buf; form feed chars, tabs and newlines are stripped away
+// - all text written is appended to buf; formfeed chars, tabs and newlines are stripped away
// - at any given time there is a (possibly empty) incomplete cell at the end
-// (the cell starts after a tab, form feed, or newline)
+// (the cell starts after a tab, formfeed, or newline)
// - cell.size is the number of bytes belonging to the cell so far
// - cell.width is text width in runes of that cell from the start of the cell to
// position pos; html tags and entities are excluded from this width if html
// Handle empty columns as if they were not present in
// the input in the first place.
DiscardEmptyColumns;
+
+ // Print a vertical bar ('|') between columns (after formatting).
+ // Discarded colums appear as zero-width columns ("||").
+ Debug;
)
// specifies the filter output. The remaining parameters control the formatting:
//
// cellwidth minimal cell width
-// padding additional cell padding
+// padding cell padding added to cell before computing its width
// padchar ASCII char used for padding
-// if padchar == '\t', the Writer will assume that the
-// width of a '\t' in the formatted output is cellwidth,
-// and cells are left-aligned independent of align_left
-// (for correct-looking results, cellwidth must correspond
-// to the tab width in the viewer displaying the result)
+// if padchar == '\t', the Writer will assume that the
+// width of a '\t' in the formatted output is cellwidth,
+// and cells are left-aligned independent of align_left
+// (for correct-looking results, cellwidth must correspond
+// to the tab width in the viewer displaying the result)
// flags formatting control
//
+// To format in tab-separated columns with a tab stop of 8:
+// b.Init(w, 8, 1, '\t', 0);
+//
+// To format in space-separated columns with at least 4 spaces between columns:
+// b.Init(w, 1, 4, ' ', 0);
+//
func (b *Writer) Init(output io.Writer, cellwidth, padding int, padchar byte, flags uint) *Writer {
if cellwidth < 0 {
panic("negative cellwidth");
}
+var vbar = []byte{'|'};
+
func (b *Writer) writeLines(pos0 int, line0, line1 int) (pos int, err os.Error) {
pos = pos0;
for i := line0; i < line1; i++ {
for j := 0; j < line.Len(); j++ {
c := line.At(j).(cell);
+ if j > 0 && b.flags&Debug != 0 {
+ if err = b.write0(vbar); err != nil {
+ return;
+ }
+ }
switch {
default: // align left
""
},
+ entry{
+ "1 debug",
+ 8, 1, '.', Debug,
+ "",
+ ""
+ },
+
entry{
"2",
8, 1, '.', 0,
"*.......*"
},
+ entry{
+ "5c debug",
+ 8, 1, '.', Debug,
+ "*\t*\t",
+ "*.......|*"
+ },
+
entry{
"5d",
8, 1, '.', AlignRight,
"g) f<o..<b>bar</b>..... non-terminated entity &"
},
+ entry{
+ "7g debug",
+ 8, 1, '.', FilterHTML | Debug,
+ "g) f<o\t<b>bar</b>\t non-terminated entity &",
+ "g) f<o..|<b>bar</b>.....| non-terminated entity &"
+ },
+
entry{
"8",
8, 1, '*', 0,
"11222333344444\n"
},
+ entry{
+ "9c debug",
+ 0, 0, '.', Debug,
+ "1\t2\t3\t4\f" // \f causes a newline and flush
+ "11\t222\t3333\t44444\n",
+
+ "1|2|3|4\n"
+ "11|222|3333|44444\n"
+ },
+
entry{
"10a",
5, 0, '.', 0,
" .0 1.2 44.4 -13.3"
},
+ entry{
+ "14 debug",
+ 0, 2, ' ', AlignRight | Debug,
+ ".0\t.3\t2.4\t-5.1\t\n"
+ "23.0\t12345678.9\t2.4\t-989.4\t\n"
+ "5.1\t12.0\t2.4\t-7.0\t\n"
+ ".0\t0.0\t332.0\t8908.0\t\n"
+ ".0\t-.3\t456.4\t22.1\t\n"
+ ".0\t1.2\t44.4\t-13.3\t\t",
+
+ " .0| .3| 2.4| -5.1|\n"
+ " 23.0| 12345678.9| 2.4| -989.4|\n"
+ " 5.1| 12.0| 2.4| -7.0|\n"
+ " .0| 0.0| 332.0| 8908.0|\n"
+ " .0| -.3| 456.4| 22.1|\n"
+ " .0| 1.2| 44.4| -13.3|"
+ },
+
entry{
"15a",
4, 0, '.', 0,
"a\tb\tc\td\te\n"
},
+ entry{
+ "16b debug",
+ 100, 0, '\t', DiscardEmptyColumns | Debug,
+ "a\vb\v\vd\n"
+ "a\vb\v\vd\ve\n"
+ "a\n"
+ "a\vb\vc\vd\n"
+ "a\vb\vc\vd\ve\n",
+
+ "a\t|b\t||d\n"
+ "a\t|b\t||d\t|e\n"
+ "a\n"
+ "a\t|b\t|c\t|d\n"
+ "a\t|b\t|c\t|d\t|e\n"
+ },
+
entry{
"16c",
100, 0, '\t', DiscardEmptyColumns,
"a\tb\tc\td\n"
"a\tb\tc\td\te\n"
},
+
+ entry{
+ "16c debug",
+ 100, 0, '\t', DiscardEmptyColumns | Debug,
+ "a\tb\t\td\n" // hard tabs - do not discard column
+ "a\tb\t\td\te\n"
+ "a\n"
+ "a\tb\tc\td\n"
+ "a\tb\tc\td\te\n",
+
+ "a\t|b\t|\t|d\n"
+ "a\t|b\t|\t|d\t|e\n"
+ "a\n"
+ "a\t|b\t|c\t|d\n"
+ "a\t|b\t|c\t|d\t|e\n"
+ },
}