Save Formatted Texts as HTML

Overview

Whilst TextEdit understands RTF and DOC nicely, it does not support HTML that well: HTML texts can be opened and saved to, but you cannot save a RTFD into HTML with image links.

OCSmart Hacks fills the gap by adding a "Save As HTML" command into any application, which uses the standard text object with a formatted text and allows the user to edit it. (For plain RTF without images you would probably want rather to use the standard TextEdit HTML save support: the Hacks one comes from the days TextEdit did not save HTML at all, and is not as flexible as the current TextEdit.)

User Interface

Find the command "Save As HTML..." in the menu: normally it would be just beside the standard "Save As..." command, but in some applications it may be moved into a different place, including a separate "OCSmart Hacks" menu–it depends on what commands and menus the application has:

Select the command. If the active window happens to contain a text object whose contents you normally would not save–like the text fields in the Find panel –, OCSmart Hacks asks whether you really want to:

Presumed you answered Yes, or the current text object contains a normal document, a standard Save As sheet is shown. It allows you to select the destination:

Note that there is an accessory view installed into the panel, which allows you to select a preset (or to open the Preferences to change presets using the "Presets..." button); the presets control the way HTML is generated, and we shall discuss them in the "Customization" chapter below.

When "Save" is clicked, the document contents is saved into HTML. Note that if the document contains images or other attachments, the selected name would be used as a folder name; the generated HTML and all the attachments would be stored into the folder. In case the document does not contain any attachments, the selected name would be used directly for the generated HTML file. Note also that for this reason the folder of the former case would have a ".html" suffix.

Customization

The conversion from the formatted text into HTML is controlled by presets. To set and edit them, open the OCSmart Hacks window selecting the "OCSmart Hacks..." item from the application menu, and choose its "HTML" tab. The list of current presets is displayed in the lower part of the window:

Use the buttons below to add, duplicate, edit, and remove presets. The presets are stored on disk (namely, in the ~/Library/Application Support/cz.ocs.OCSAttributedStringHTMLExtension/Presets folder): if there is a possibility they were changed manually or by another application, use the "Rescan" button to read them in again.

A preset is a plain list of attributes, which the HTML generator uses. At this release, the attribute list is quite short: clicking the "Edit..." button you can open a sheet, which contains a preset editor:

The preset name and comment you can use to keep presets organized. The meanings of the other widgets are following:

Full Page: if checked, a complete HTML page, including the <!DOCTYPE>, <HTML>, and <HEAD> tags, is generated. If unchecked, only the contents of the <BODY> tag is generated: a partial HTML document, which can be easily added into another one;
Title: the text field, enabled for full page documents only, sets the contents of the mandatory <TITLE> tag;
String Encoding: select any string encoding you want to use for the generated HTML document. In the full-page mode, the encoding will be properly emitted into a <META CHARSET> tag. The generator would export properly any character; those which cannot be expressed in the selected encoding will be converted into Unicode entities (&#<unichar number>;). Also note that the UTF-16 encoding will not generate a UTF-16 HTML (which would be unreadable in a vast majority of browsers), but a plain ASCII HTML in which all special characters are expressed using the Unicode entities;
Folder name of Generic attachments: the name of the folder into which any non-image attachment will be stored. The generated HTML will contain an <A HREF> tag to link the file. If empty, the attachments would be stored beside the generated HTML file;
Attachment link template: the form of the link to an attachment. It can contain "%@", which will be replaced by the original attachment file name;
Folder name of Image attachments: the name of the folder into which any image attachment will be stored. The generated HTML will contain a <IMG SRC> tag to show the image. If empty, the attachments would be stored beside the generic ones;
Downsample: resize all images to N%;
After downsampling, constrain image to max: if the image (after downsampling) is still wider than the number of pixels set here, it will be further shrunk to fit the width;
Type Parse and Generation rules: the actual rules which drive the conversion are at the moment not editable in the GUI. A power user can set them up using any XML editor, since presets are standard XML files; this, along with the rules themselves, is described in the Power User Help page.

Default Rules

The default set of rules converts RTF(D) to HTML this way:

a link in the RTF(D) is converted to <A HREF="link contents">;
if the font happens to have a nonzero italicAngle, tags ... are used;
if the font name happens to contain "Bold", tags ... are used;
if the font pointSize is greater than or equal to 24pt, tags ... are used;
otherwise, if the font pointSize is greater than or equal to 18pt, tags ... are used;
if the font pointSize is smaller than or equal to 12pt, tags ... are used;
if a paragraph has its firstLineHeadIndent less than its headIndent, and if it begins by a bullet or a hyphen, followed by a space of a tab, tags <UL>...</UL> are used (along with the appropriate <LI>...</LI> for consecutive ones);
if a paragraph begins in a fixed pitch font*, tags <PRE>...</PRE> are used;
if a fixed pitch font* is used elsewhere, tags <TT>...</TT> are used;
each paragraph is enclosed by ... (unless it is inside <PRE>);
embedded images and other objects are converted as described above.

*A font is considered fixed pitch if its 'i' and 'm' are of the same width.