The HTML-form creation dialog box writes one or several files formatted in HTML (= hypertext markup language, the format used in the word wide web). This dialog box can be called from the DeltaAccess Switchboard or by opening the form entitled ToHTMLForm in the database explorer.

You can use HTML forms as an off-line editor to add new items or edit existing items over the world-wide-web. See the following chapter for information about re-importing the information you will receive back.

The HTML form generation dialog contains a multitude of options to customize the HTML code generated. Naturally, the amount of options can be confusing. Please note that the only required options are found on the first two tabs, General and Output files. You must select a descriptor project and specify the post method and address to which the edited data shall be returned. With these minimum settings, you can already generate a form for new items. Beyond that, the three most useful things you can do to optimize the output are:

• Set the project properties (ProjectAuthors, ProjectEditors, ProjectCopyright, ProjectTitle, and ProjectIcon) which are used in the default header of the forms. See Editing project properties for further information about how to view and edit project properties.
• Revise your character definition and set the MultistateType to Exclusive (= 0) for all characters which are not really multistate, i. e. where only a single state may be present in any item. If you are importing DELTA files, you achieve the same effect by using the appropriate character types EUM/EOM for these characters.
• If you have numerical characters where a range or minimum/maximum is never expected (e. g. "number of species in genus") you should revise the character definition within DeltaAccess, deleting the unwanted numerical states. Consult the paragraph Numerical attributes at the end of Output options for further information.

The number of parameters available (more than 80 parameters can be set) makes it likely that some combination may not work as expected. Please contact me in this case, providing a documentation of your settings, an example of an HTML form which looks unexpected, and either a description of what you expected, or a manually edited HTML form (perhaps one character only) which confirms to your expectations. Please try the default settings first and modify them only if you find them unsatisfactorily.

The following sections discuss the options on each tab page:

   General
   Output files
   Index file
   Include files
   Output options
   Formatting options
   Font formatting options
   Localization strings
See also
   Example

You can not use or change a user-defined sequence of character states in an item description, which is different from the default sequence defined in the character definition. In DELTA and in DeltaAccess you can specify for a certain item and character, that states "2 or 1 (rarely)" are present. After using an HTML form to update the date, this will be changed to "1 (rarely) or 2".

I may be able to provide additional options for customization. If you want a different appearance for you forms, please modify a form manually, and explain separately what you have done. If this can be done automatically based on the information in a DELTA character definition, I will try to include it in the next version.

Return to Export    Main table of contents

On the first tab page you must select a project name. For each descriptor project (including subset projects) you can save all other settings in this dialog to retrieve them later, using the Save settings and Open saved settings buttons placed directly under the project name field. The settings are stored as a property entry in the PROPERTY table. Note that it is possible to "copy" the settings from one project to another: Select the project for which settings have been previously stored, click on Open saved settings, select the project for which you want to use the settings and click on Save settings.

The option Form, post action at the bottom of the tab page determines how you want to retrieve information sent back to you over the World Wide Web. If a user presses the Send item description button in her or his web browser, the browser will send the contents of all elements in the form over the internet using a "post" action. This information will be sent to the URL that you entered in the field at the bottom of this tab page.

The simplest way to retrieve the information is to enter a mailto-action in this field. This is especially useful if you do not have direct access to a web server, which would have to react to other types of URLs. You will receive an email under the address you enter here, containing the returned data in an encoded format. See the following chapter for information about re-importing the information you will receive back. Example: If your email address were yourname@yourserver.edu, you would enter 'mailto:yourname@yourserver.edu' (without the quotation marks) in the post action field.

If you want to program a web server to directly react to the action, you can enter a full URL, or simply a path. If you enter /Delta/response/ the web browser would sent the reply to this folder on the same server from which the form was received by the browser.

Return to HTML form overview    Main table of contents

On this tab page you can select which HTML documents the procedure will generate, and how they shall be named. Available options are

Empty form for a new item: Create an empty form directly from the character definition. This form can be used to add completely new items to a descriptor project over the internet.

Forms for each existing item: Create a similar form for each existing item in the current descriptor project. The fields of the form will already contain the information that is present in the database. Thus, text fields will be filled, and checkboxes or radio buttons appropriately set.

Return to HTML form overview    Main table of contents

The options on this page are only available if the option Forms for each existing item on the previous tab page has been selected.

Additional index file index.html: To help with the generation of a web site with many item description forms, a basic HTML index file is generated. It contains hyperlinks to all HTML forms generated. The index file contains the same header as the item forms, which may not always be appropriate. It is assumed that the index will generally be manually incorporated into a more complex HTML document. The index file is always named Index.html. You can rename it if you want to support HTML forms from several descriptor projects in a single folder.

Sort index by: You can further choose in which order the links to the item forms are included in the index file. This also influences how the entries in the index are written. If you choose IID, the entries will start with the IID followed by the item name, if you choose item name, the item name comes first followed by the IID in brackets. Note that for technical reasons, you can not sort the index by IID if you derive the filenames from item names (= third option in the filenames option group, see previous tab page).

Internal index: The entries on the right side concern an additional internal index within the index file. This internal index helps to jump directly within the list of item links to all items starting with a certain letter. An internal index is especially useful if the index contains several hundred items. This option is meaningful and available only if the index is sorted by item names (see previous option).

Item names	First word index	Second word index
Anathema excelsior Anathema erasii Anathema xanthippii Bertholdia arnoldii Bertholdia borriquino Deltaria janthii	A B D (Note that 'C' is present neither in the item names nor in the index)	Anathema E X Bertholdia A B Deltaria J

The option labeled Link prefix may contain the publishing folder for the item descriptions. The text in this field becomes part of the hypertext link in the index file. If it is left empty, most web browser will work well as long as the file is placed in the same directory as the index file. If the prefix does not end with a slash ('/'), a slash will be appended. An example for a valid link prefix would be: http://www.bgbm.fu-berlin.de /Projects/DeltaAccess/

Return to HTML form overview    Main table of contents

This tab page allows to define two include files, which will be included before and after the form part in each HTML document created.

The include files should contain text or HTML codes which can be inserted directly into the HTML files generated by this procedure. The text read from the include files will not be modified in any way, thus you can use any HTML formatting you wish to. On the other hand, you must already transcribe things like accented characters into their HTML transcriptions (e. g. use "ü" for "ü").

The include files may not be a full HTML files though. They may not contain the usual <HTML></HTML> start and end codes, and they may not include any header information. To produce include files, it is easiest to modify an existing form in any HTML editor, save the file, and open it with a plain text editor. Then copy the start and end part, which shall be inserted into the new HTML forms, into separate files.

Normally, DeltaAccess will produce a default HTML header, which includes information read from the project properties ProjectAuthors, ProjectCopyright, ProjectTitle, and ProjectIcon (see Editing project properties for further information about how to view and edit project properties). You also have the option to customize the HTML color definitions used in this default header. The color definitions define which colors are used in the web browser for text, background, and links.

Alternatively, you can select the option to Replace the default start with the include file for the start. In this case, in contrast to what was said above, the start file must include all necessary HTML start and header tags. Still, it may not include an </HTML> tag indicating the end of the HTML file.

Return to HTML form overview    Main table of contents

This tab page offers the option to define the information included in the HTML forms generated. Except perhaps for the option to include headings (which improves the structure and usability of the item descriptions), the options on this page increase the complexity of the forms and make them more difficult to use. Choosing the right balance between easy-to-use forms and including all information is often difficult. The default settings will export all available information, but do not, for example, allow entering notes for a newly selected character state. If you want to use the forms for true revision over the internet, you should not restrict the output further, or you may loose information during the revision. On the other hand, if you mainly want to present your data, and use the forms to get a more general feedback from your colleagues, you can disable special character states (like U/V/-) and even omit all modifiers or notes.

Include headings from character definition: If you have added headings to your character definition and defined them as CharHeading or ItemSubheading, you can improve the structure of the item descriptions form by inserting the appropriate headings in front of the character group to which they apply. See the chapter about the heading editor (EditHeadings) for further information about headings.

Internal navigation within the HTML document using headings: If add internal index of headings at top is checked, a bookmark will be added to each heading and a hyperlink index for the heading bookmarks will be added at the start of each HTML form document. This enables the user to jump directly to a specific heading instead of scrolling down to find the relevant information. The "'Go-to'-phrase for internal index" on the tab page Localization strings 2 is used in front of this hyperlink index. The checkbox is disabled if no headings are included. If add link back to index after each head. is checked as well, a hyperlink (text or image) is inserted after each heading, to return to the index of headings at the top of the HTML document. The "'Back-to-index'-phrase or image" on the tab page Localization strings 2 determines the text of the hyperlink (default: "-- back to index"). If this phrase ends with '.gif', '.jpg', or '.png' it is assumed to be a relative or absolute URL to an image that you supply. Please contact me if you need to use other file extensions than the 3 mentioned above.

Special character state options: Special character states (U/V/-/TE, for Unknown, Variable, Not applicable, or Special text, respectively) are not included in a DELTA import file. Since they are implicit in the definition of the DELTA format, DeltaAccess adds them automatically during import. The default settings of the HTML form generation dialog box ensures that all existing special character states are included in the item description output. The option Autodetect ensures, for each character separately, that whenever a special character state has been used in at least one item, it will be included in the form for all items. This setting will usually work well with descriptor projects that already contain many items. An exception is the special character state 'TE' in categorical or numerical characters. Since it must either be already used in an imported data set or manually added to the character definition, it is output wherever it has been defined (option Always). You can also suppress the use of special states (option Never). Please note, that suppressing information will cause problems if you want to use the form for the revision of existing items over the web, since you will loose the information encoded by special character states.

Alternatively, you can check in which characters a certain special state has not been used and will never be used. For example, you may not want to use the variable state at all, or the definition of character dependency may be sufficiently well developed that the explicit inapplicable state is never needed. In this case, you can remove these states from the character state list in the character definition; see the character states section of the character definition editor (CharEdit). Special character states removed from the character definition are not included even when the option Always is used.

Numerical attributes (Min, Max, Ranges, etc.): Sometimes, a minimum and a maximum may not be appropriate for a real or integer numeric character. Currently no option is available to skip some numerical attributes for a subset of the numerical characters. A future version of DeltaAccess could use an Autodetect mechanism similar to the one used for special character states (see above). Nevertheless, you can eliminate the inclusion of Min, Max, or Ranges in the HTML form. Since an imported DELTA definition lacks any information about this, DeltaAccess adds these attributes during import. It does not require them to be present though. Within DeltaAccess, you can define exactly which numerical attributes you want to be preset. You can simply delete the unwanted numerical attributes in the character states section of the character definition editor (CharEdit). Note that if Mean, Median, Mode, or Value are left as the single numerical attribute, the HTML form will include it without the usual table display and without a label. It is assumed that the character name gives sufficient explanation in this case.

Include item descriptions … Notes (= comments) and Modifiers:

At the bottom of the output options tab page you can determine the rules for adding item description Notes and Modifier attributes to characters or character states. Notes are equivalent to the standard DELTA comments (which are no true comments, but data items). Modifiers are a special feature of DeltaAccess. During import of a DELTA file, many comments are recognized as modifiers and removed from the notes. See the chapter Introduction to the concept of modifiers for further information.

Adding a text field for inserting the modifier and the note for each character state makes the resulting forms look rather chaotic. Most of these fields will not be used, since usually only 1 or 2 character states are active, and you can add notes only to activated checkboxes (note that this can not be enforced in the HTML form itself!).

The default settings are therefore to output only existing modifiers and notes. This allows a complete revision of the data for existing items, but no new notes or modifiers can be entered. As an alternative to adding a text box for notes for each character, you can select the option existing, plus single new. This will output an editable text field for any note which is already present in the data set. In addition, it will add one further text field to enter a new note. The state to which the new note applies can be selected from a pick list of all available states.

The settings can be set separately for categorical (type UM/OM), numerical (type RN/IN), and text (type TE) characters. Modifiers are not defined for text characters, and since only a single Note can be added to a text character, the option existing, plus single new is not available here.

Return to HTML form overview    Main table of contents

This tab page contains many options to customize the appearance of the forms generated. No setting on this tab page will influence the data included in the form, while settings on the previous Output options tab page influence both the formatting and may suppress existing data.

The first option influences, whether characters will be separated by horizontal lines or not. The default is Not consecutive text/numeric. It means that the horizontal line is skipped between two consecutive text characters (type TE), or two consecutive characters of the same numeric type (RN/IN). Note that lines will be inserted between text and numeric, as well as between real numeric and integer numeric characters. If you select not within LinkGroup, horizontal lines will be suppressed within the LinkGroups defined in the character definition. All characters not in a link group will be separated by a horizontal line. You can edit the link groups in a special editor, see Editing link groups.

The second option is used to enable text boxes with multiple lines and scrollbars. By default, all text characters are output as input boxes, which display only a single line. HTML forms offer multiple line text windows, which include scrollbars (using the <TEXTAREA> HTML tag). You either always use these text boxes, or you can use them only for those characters, which contain a larger amount of text. Enter the character numbers of the desired text characters into the text field below the radio buttons, leaving a blank between the numbers. The button with the lightning symbol () will automatically select appropriate text characters. It will include all text characters, where at least in one item description the text is 1.5 times longer than the current width defined for a single line text input field. Click on the -button to change the list of text characters manually.

At the right side, immediately above the radio buttons, you can also change the height of these text boxes in lines. Only text characters can be changed to windowed text boxes; the special text state 'TE' possible for categorical (UM/OM) characters is always treated as single line input field. The HTML text boxes use the tag <TEXTAREA WRAP=SOFT>. This causes lines that are wider than the display area to wrap to a new line. Hard returns inserted by the user are preserved and transmitted back.

The next section is concerned with the width of the available editing elements. The first two define the width of single line normal text, and the multiple-line text boxes. Note that multiple-line text boxes usually appear wider than single line text input boxes, which is why the defaults are set to 70 and 80, respectively. Special text is used whenever the special 'TE' character state has been enabled for a categorical character. This feature of DELTA can be used to allow the user to enter information, which is not available in the list of categorical characters. The phrase used in front of these special text characters can be set on the tab Localization strings. Numeric input is used to enter the values for numeric characters; Modifier text and Notes text are used for the width both of directly output modifiers/notes, as well as for the combination of an additional modifier/note with a character state combo box. Note that modifiers and notes for those numerical attributes that are presented in a table (Min, Max, Mean, etc.) always use the numeric input width as well, to keep the table properly formatted.

The option Observe NewParagraph attribute will insert an additional paragraph in front of each character where the NewParagraph has been set in the character definition.

The option Start each char. state on a new line will start a new line in front of each character state checkbox or radio button. Instead of being displayed as continuous text, wrapping to the width of the web browser window, the states will now be displayed in a tabular form. As long as the browser window is wider than the longest character state phrase, this format may be easier to edit.

The option Observe MultistateType = Exclusive... allows the use of radio buttons or a pick list instead of checkboxes wherever the character is defined as 'exclusive'. These controls are more appropriate than checkboxes for this purpose, since you can only select a single option. If you have many character states, giving all options as check boxes or radio buttons takes up a lot of space and can be confusing to the user. A pick list (= combo box, HTML SELECT tag) is smaller and easier to use in this case. Directly beneath the checkbox, you can set up to how many states radio buttons will be used, and from which number of states onwards a pick list will be used. If you enter a 1, all exclusive characters will be output as a pick list; if you enter a large number like '999' all exclusive characters will be output as radio buttons. By default up to 4 states are output as radio buttons. Note that due to the inclusion or exclusion of special states (U/V/-/TE), the number of states is not necessarily identical with the number of states in the character definition. The number entered here refers to the actual number of radio buttons that will appear on the HTML form. See the Include U, Include V options on the tab page Output options for ways to influence the availability of special states.

Return to HTML form overview    Main table of contents

This tab page contains additional formatting options, mainly the character formats (font name, size, color, bold or italic font).

In addition to the fonts formats itself, the outline level of several parts of the form can be set. An outline level of 0 indicates body text, levels 1 to 4 indicate the first to 4th heading level. If, e. g., character name is set to outline level 3, it will be marked with the HTML tag "<H3>" and displayed with the font, size, and color which was defined in the web browser for this outline level.

Note that too much formatting results in lengthy HTML documents, which may not look identical on all web browsers or may be difficult to read at some screen resolutions. The latter is especially true for the settings of size, which is normally set by the user specifically for her or his web browser. Leaving size empty and font name and color on default settings will preserve the default appearance of HTML documents and will not generate any HTML tags that increase the size of the HTML documents.

Return to HTML form overview    Main table of contents

All phrases used by the program in the forms generated can be modified on the two tab pages labeled Localization strings 1 and Localization strings 2. This is useful for general customization as well as for the generation of forms in languages other than English.

Note that some formatting features can simply be disabled, by removing the string on the localization tab. For example, you may not want mandatory characters or explicit states to be marked in any way. To achieve this, simply delete the respective localization phrase.

Label for special user code/email address: This adds a field directly above the send/reset buttons of the HTML form, asking for either a user name or the full email address of the person editing the data. When you use a CGI-gateway instead of the 'mailto'-method to receive updates over the internet, you no longer have an email address of the person sending you updates. Some means of verification is usually required. If you use a CGI-script, it can directly reject submissions where this information is missing and inform the user accordingly. Simply remove the label if you do not want this feature.

For modifiers and notes, you can define a start and end phrase. The defaults are simple parenthesis, but you could also use a label like "Modifier:".

The Phrase for add single new and Character state pick list for add single new ... apply only if you have selected the option Existing, plus single new for Notes or Modifiers on the Output options tab page. The sequence of phrases and controls used for these options is (using Notes as example): Notes start [e.g."("]; Character state pick list phrase start; <Pick list to select a character state>; Phrase for add single new note; <Text field for the Notes themselves>; Character state pick list phrase end; Notes end [e.g.")"]. The checkboxes New line after the Phrase for add single new apply only to the single new modifier or note as well. If checked, a new line will be started before the whole add single new sequence.

Return to HTML form overview    Main table of contents

The following excerpt of a DELTA character definition was the basis for the parts of an HTML form shown at the end ("[…]" indicates omissions of part of the definition).

[…]

[…]

[…]

Return to HTML form overview    Main table of contents

Two methods exist to receive the revision information entered by a user on the internet: email and cgi-gateway. This method is determined when the forms are created on the General tab page of the HTML-form creation dialog box.

• CGI-gateway: This method requires access to an internet web server, on which a small program (often written in Perl) has to be installed. A single program can be used for all your descriptor projects. An example Perl script is currently developed and tested in cooperation with Gerhard Rambold and Loran Pezdirc. This script saves the post information it receives to a file to be further processed by DeltaAccess.
• Email: This is the simplest way and does not require access to an internet web server. If you are using the mailto:<your_email_address> setup for the form, post action , you will receive an email whenever somebody has sent a revision to you. You should save this email to a file. If possible, save it directly without opening it first, since some email readers will reformat the data during opening, and will save the reformatted data. See Saving posts in common email programs.

In the re-importing form, select the saved email file under 'File containing post:'. If you saved your files with an extension '.dat' they will be directly displayed in the dialog box that appears when you click on the file-open button on the right side; otherwise you must change the file type at the bottom of the file-open dialog box. You may use wildcards (* for any number of letters, ? for a single character) in the filename to import multiple files with a single command. For example, importing 'C:\Data\S*.dat' will try to import any file in the named directory which starts with 'S' and has a file extension 'DAT'.

Please note that, although a project name can be set on the import dialog box, you can leave the field empty if you want to re-import data into the same descriptor project from which the HMTL forms were originally produced. The original project name is part of the data returned to you. You can import data into another descriptor project, but it is your responsibility that the character definitions of the two projects are compatible. The data in the post contain the character ID (CID), so for each CID the character type and available character states must be identical with the original project. Other changes, e.g. in the character or character state name, notes, mandatory definition, etc. provide no problems.

See Re-importing data from HTML forms, Import options for information about the options available on the second tab page of the dialog box.

Click the button 'Re-import HTML form postings' at the bottom of the dialog to perform the import.

If the import process encounters any problems during import, a DELTA coded text containing the information about the current item is displayed. You may use it to identify the problem in the DELTA coded text and use this manually in an DELTA import. Please note that depending on the type of error, the import may have been partially successful and some data may have been added to your descriptor project.

Return to Export    Main table of contents

Saving posts in common email programs

Re-importing data from HTML forms    Return to Export    Main table of contents

Currently available options are:

Re-imported items contain the ID of the item they were created from. This information can be used to:

• Append as new item under new IID. The old item remains unchanged under its original item ID (IID) and the revised item is added as a new item under a new IID.
• Directly overwrite original. This is a dangerous option, since you may loose your original data if the re-imported data are incorrect, or even if the re-import fails. If the old IID is not available, or can not be deleted (e.g. because some records are currently edited and therefore locked), the re-import will create a new item instead like in the first option.
• Directly overwrite, copy old item to a backup project. Before the item is overwritten, the item definition and description are copied to a backup project. This is the recommended option and it is used by default. The project will remain to function as if the revision had been made in a local item editor, but you can still trace the changes made over the internet. A backup project is a project with independent item tables, linked to the character definition of the base project (i.e. a mixed project) and with the name of the base project plus the suffix '_IBK' (for item backup). Example: If the base project name is 'LIAS', the name of the backup project will be 'LIAS_IBK'. The backup project will be automatically created if it does not exist yet.

Overwriting an item means that all item descriptions are deleted and the new descriptions are then imported. However, the item definition itself is not deleted and then reinserted, but directly modified. This means, that additional information linked to an item (including item notes, or resources) remain intact.

The overwrite options do not apply to new item forms. Data from such forms are always appended (they use IID=0, which is recognized by the re-import procedure).

The purpose of annotations is to save information about the original item, which was revised over the internet, its possible backup location, the name of the person responsible for the revision, and the date of revision. The following options are available:

• Do not add. No annotation is stored at all.
• Add to item note. The annotation is appended to the text already present in the ItemNotes.
• Add to a text character. Like the option above, only that the information is added to a specified text character. The character for annotations can be set using the provided button. The value is stored in a project property separately for each project. Thus, if you plan to import data from multiple projects, do the following: Select each project on the first tab page of the dialog box, click on the Select CID button on the second tab page, and select the character. After you have finished doing this for each project delete the project name on the first tab page; otherwise data from all projects would be imported into that project.

• Show DELTA code after import: by default, the DELTA code for the imported item is shown only if an error occurred during import. If you activate this option, the code from all imported items will be displayed in a message dialog box. You can select it and copy it to the clipboard (Ctrl-C) to use the code elsewhere.
• No warning if project names differ: The postings which you import contain the project name of the project from which they were originally created. You can re-import them into any compatible project, but by default a warning is issued for each imported item. Activating this option suppresses the warnings.
• 'Item comparison' after each import: This option automatically opens the item comparison form (*** NEW; NOT YET DOCUMENTED) after each item that is imported. The form allows to check the differences between the original version of the item and the revision sent in over the internet. The form can and will not be opened if either the item was directly overwritten without any backup, or if a new item was imported. If several items are imported from a single file, or if several files are imported using wildcards, the import process will continue after the item comparison form has been closed.

• The currently available test version of an item comparison is not yet finished, a more sophisticated summarize/contrast option is planned which allows writing out the contrasts or summarized data.
• The import process has been written to deal with multiple items (posts from HTML forms) in a single file. You may want to try it, but please be warned that this feature has NOT been tested well.
• Should the header information of the email (containing the sender, date, etc.) be added to the new items?
• The item comparison form allows updates only of item name, notes, etc., but no changes in the item description. This is due to technical limitations of the database engine and will be difficult to overcome in the current version of Microsoft Access.

Re-importing data from HTML forms    Return to Export    Main table of contents    Next