| Next | Contents |
leaf.table.Endgrid – should flush all text, as it might be buffered across more than 1 page wide and unflushed buffers can crash out later on. Fixed.
Viewer – 16-bit ŒNA switches removed as they are no longer required and failed under Dyalog Version 10. Assume 32bit only from now on.
table.Spread – when using <b> in a table AND at the same time specifying the same cell to have bold emphasis, the cell gets an incorrect font. Fixed.
Interim Release as 1.8d on 01/05/2003
HTML rendering – new property added to text, ruler and bitmap objects on the master page layout to allow these to be selectively included on reports rendered as HTML. By default, no items from the page layout are rendered to HTML, as the assumption is that the newleaf report will form part of a page design which already has some kind of framing pre-built.
leaf.Include – was not flushing bookmarks (also leaf.IncludeLeft). Fixed.
Wrapped text on master page – accepts any of ŒAV[4],'||' as hard line-breaks, either in the ‘boilerplate’ part or in a substitution string. See Subtest for the updated example.
leaf.table.List – measured cells (with headings) which wrapped over several pages could create Length Errors. Fixed.
leaf.Richtext – ignored wraparound block set up by BitmapLeft. Fixed.
leaf.DefineFont – could be confused by re-using a ‘standard’ name such as ‘AR’ for the new font. Appeared to work, but picked up the wrong metrics so wrapping messed up. Now removes the matching entry from the metrics cache.
leaf.Reset, leaf.table.Reset – set paragraph / table properties back to defaults. Probably safer than resetting explicitly if many properties have been used in a previous table.
Bookmarks – were incorrect if leaf.SkipToPage had been used to update the page-number in flight. The code to find the bookmarks in a spooled report assumed it could use the page# as the file index. Fixed.
Frames – pattern set extended to include semi-transparent fills and simple gradients (gradients work well in PDF). Also a frame with shading but no edge was not being shaded. Fixed.
Tables – List and Spread were slightly inconsistent on page boundaries. A table which exactly fits the framesize should always be OK, but List threw a new page here. Also some code in Spread to tidy the result of merged cells could occasionally cause the last row of a split table to be extended downward to the edge of the frame. Both issues fixed.
Bitmaps – records file info along with boundingbox. This saves time when converting to PDF format as we don’t need to go and find the real imagesize all over again.
leaf.Group – this was getting too hard to extend, and is mostly superseded by table headings and captions, so this one is not longer supported.
HTML support – may now be generated like any other output type from the spool file or a PG variable in the workspace. To enable this, you must run leaf.Mirror 1 to enable mirrored output in PG, as this format requires more details of the original data to be retained, particularly for complex tables.
myhtm„html.PS PG © Default style myhtm„CSS1 html.PS PG © Sample stylesheet to make it prettier
CSV support – uses the mirrored content in the same way as HTML. Ignores any output from Place and Flow, so the only content in the result comes from tables (List or Spread). Table headings, subheads, captions and footnotes are included if they have been specified.
mycsv„csv.PS PG © No other options here
Bitmaps – enhanced behaviour is now as follows. If we are at the very top of a frame, the picture is scaled to whichever is less of height and width. If we have moved down the frame, then is is constrained to the width only, and we will skip to a new frame (or spill out of the bottom) depending on the overflow setting. The same applies to BitmapLeft if the picture does not fit.
Overflow property – this is now upgraded from a boolean switch to have 3 possible behaviours when a frame is filled:
If you select the ‘Discard’ setting, output will be thrown away until the next call to Newpage or NewFrame, either explicitly or implicitly by a command such as Switch which resets the page layout. See SkipTest for examples.
Page skipping – paragraph text was allowing for ‘space after’ when checking to see if there was enough room without leaving an orphan. It should just look to see if there is room below the last baseline for the descenders. Also we turn off all widow/orphan checking if overflow=2 as we will never see any of the following text, so we may as well use all the frame here!