This section is edited by Dott.ssa Susanne Müller and Dott.ssa Costanza Giannaccini for contents and structural aspects, by Dott.ssa Francesca Di Donato and Dott.ssa Chiara Aiola for the formal syntax.

1. Introduction
2. Letter structure

2.1 General indications
2.2. Semantic Units
2.3 Opener
2.4 Text body
2.5 Closer
2.6 Other
2.6.1 Page numbers added by the author and conventional signs for page endings
2.6.2 Attachments

1. Specific cases

3.1 Abbreviation Brevigraph
3.2 Closing Scribble
3.3 Catchword
3.4 Handshift
3.5 Ligature
3.6 Additions
3.7 Hi (rend “XX”)
3.8 Gemination
3.9 Inverted commas
3.10 Del
3.11 Spaces
3.12 Specific terms and acronyms on page bottom
3.13 Square parenthesis together with Printed in dates
3.14 Gaps and illegible characters conveyed with the symbol “%”

1. Conversion chart
2. Semantic Edition checking procedure

1. Introduction

Purpose of this document is to clarify the essential steps that from a letter transcription in XML format lead to its transformation into the so-called Semantic Edition, used in semantic annotation, and for final users interested only in the letter contents, rather than in its philological aspects.
The main steps are the following:

1. XML freezing. Technically, this procedure consists into copying at a given point the XML version of the transcription from the field n7tra_xml_content (used for the philological visualisation of the letter) on the field testo_costituito_xml
2. Transformation of XML text into HTML via an appropriate XSLT stylesheet (profile OXGARAGE: burckhardt_testo_costituito)
3. Visualisation of the HTML in the backend (in Pundit or not) with the appropriate CSS dedicated

2. Letter structure

2.1 General indications

Paragraphs: All paragraphs are left aligned. All div tags with

must be left aligned, paragraph attributes (eg. rend = ‘indented’) are not considered in the visualisation of the Semantic Edition.

Linebreak: No linebreak should be visible. Tags at the beginning of each line are not considered as markers of new paragraphs. Even sequences (eg.: ) used to indicate a blank line between letter parts are not considered new paragraphs.

2.2 Semantic Units

It is considered “semantic unit” a letter portion marked with a specific tag for which is agreed the use of standard spacing, regardless of specific new paragraphs. Within a single semantic unit the text is kept on a single line. No
are inserted, but any new paragraph is handled via css/js after conversion.

Semantic units identified by the tag:

• unit grouping date and place: http://www.tei-c.org/release/doc/tei-p5-doc/en/html/ref-dateline.html
• – date of the letter, no final punctuation, no parentheses, nor prepositions or initial articles eg. (den 11.10.1884.)
• – salutation, variable length. Only present in the Opener.
• signature, variable length. Also in this case, if a signature is split into multiple tags , each tag is a semantic unit in itself.
• place: Trieste-Zürich-Paris no final punctuation eg. (Trieste-Zürich-Paris,) http://www.tei-c.org/release/doc/tei-p5-doc/en/html/ref-dateline.html. The place related to the sender is considered inside the dateline, since it indicates the place where the letter has been written. On the other hand, the place indicating the address of the recipient is not inside the dateline. The implicit references to sites: dahier, ebenda, hier, etc. won’t receive ANY tag.
• recipient: name and address: only use and for recipient name and address; titles are kept in the recipient, but not the prepositions (An, Al etc.) (eg. ID 37, 66, 77).

A new paragraph is added before and after each semantic unit, except for place and date.
Notes: Other letter elements having own semantic meaning, but not matching a specific encoding, follow the general rules of the element where they are contained.
If semantic units are split and both reading and understanding prove to be difficult, then an editor’s intervention on the Semantic Edition will reorganise and merge the separated parts.

2.3 Opener

Here is applied the division into semantic units (each semantic unit appears on the Semantic Edition as a continuous line – no new paragraph).
If no new paragraph is present between Salute and text, nor should it be in the TC, see ID 543.
If also the dateline is missing, then the text begins with the

see ID 333, 381 and 381

2.4 Text body

Only the subdivision into spaced paragraphs is kept (XML tags

are converted to HTML tag

).

2.5 Closer

Here is applied the division into semantic units (each semantic unit appears on the Semantic Edition as a continuous line – no new paragraph).
It contains the tags , , , , .
N.B. Signed also includes indication about the letter being written “on behalf of”. eg. j-g-cottasche-buchhandlung-(muenchen;-stuttgart)-TO-burckhardt-jacob-1880-07-26 (), pettinati-nino-TO-burckhardt-jacob-1885-03-24 () and his-eduard-TO-burckhardt-jacob-1862-02-27 ().
In case of multiple signatures, these will be added one after the other. Eg. heußler-hans-TO-burckhardt-jacob-1892-09-18
In the event that a signature is on the same line of the Salute, then a new paragraph is added and the signature is inserted in the Closer (see. ID 76 and 76: Crippa).
If the sender address is added at the letter bottom, this will respect the Place rule, but in case the writer’s name is repeated, this won’t be signalled. Eg. pettinati-nino-TO-burckhardt-jacob-1885-03-24 () and clara-cohn-01.10.1890 (

2.5.1 Postscripts, whether explicitly defined as such or not, should be included in the Closer. Tag .

2.6 Other

2.6.1 Page numbers added by the author and conventional signs for page endings
They shouldn’t be displayed in the text. This occurs thanks to a specific coding identifying them.
Tags: 1
The conventional sign ./. introducing a new page should not be displayed.
Tags: ./.
The abbreviation W S G U has conceptually the same meaning, it will be therefore treated as ./. (i.e. non displayed).

2.6.2 Attachments
Only B1 type Attachments (see Protocol Metadata, Section Attachments).
Annexes appear after the letter transcription and follow the conventions described so far.

3. Specific cases

3.1 Abbreviation Brevigraph

They are displayed in their extended version (eg. Exclamation/questions marks without full stop will appear as normal punctuation marks, scribbles conveying words or syllables appear in their extended version, etc.). The version shown in the HTML is the one corresponding to the tag .

3.2 Closing Scribbles

They are final signs marking the end of a line, page etc. They just follow the text flow since they are not included in any tag.

3.3 Catchwords

The tag Kustode must be modified so to display the second occurrence (the kustode must tag the first occurrence).

3.4 Handshift

Handshifts are present in the text, but appear in a different style. In the HTML it is included in a tag.
Handshifts regarding only underlinings will NOT be visible in the Semantic Edition: the underlinging is visibile only in the Philological version with an appropriate description.

3.5 Ligature

In German letters (not in the French ones) they must be displayed, eg. Æ, æ, Œ, œ etc.

3.6 Additions

In principle, the editor inserts the additions where they are on the manuscript.

Special cases
3.6.1 Graphic sign for reference (cross, asterisk or similar): additions are inserted, if possible, where the author has placed them. Two cases can be distinguished:
a) The author adds the additional text that is part of the text flow, the phrase does not undergo any syntactic alterations: the editor inserts the additional text in the location wanted by the author, so no graphic sign is needed (even if present on the manuscript)
b) The author adds additional text that is difficult to insert in the speech flow or is syntactically not congruent: the graphic sign of reference is left where it is located on the manuscript, but the additional text is attached at the end of the letter.
3.6.2 No graphic sign for reference (cross, asterisk or similar). Two cases can be distinguished:
a) The author adds additional text that is part of the speech flow, the phrase does not undergo any syntactic alterations, but is rather completed by the addition: The editor inserts the text tacitly in the location that the editor considers as the most appropriate
b) The author adds additional text that is difficult to insert in the speech flow or is syntactically not congruent. The editor attaches the additional text at the end of the letter.
(hans-von-und-zu-aufseß-1872-04-04)
(alioth-august-maximilian-1880-11-13)
(moritz-heyne-1877-02-20)

3.7 Hi (rend=“XX”)

3.7.1. Underlinings
a) double underlinings are normalised as single (while on the Philological version they are kept as such
b) Underlinings both above and below one or more words are simplified with single underlinings (a comment is added only on the Philological Version)

3.7.2 Superscript

3.7.3 Printed: for the printed text is used a different font

3.8 Gemination

All Gemination types, whether they replace just parts or whole syllables (eg. pier-desire-pasolini-10.4.1890), or suggest the doubling of a consonant (eg. Ludwig-geiger-1884-12- 15), will appear on the text with the corresponding special sign (̄). The XML won’t be modified in order to remove the gemination coding, but it will only be displayed the .

3.9 Inverted commas

Inverted commas must be visible. In the event that inverted commas haven’t been transcribed with the graphic sign but included in the encoding (eg. Ohr), they will be explained through the sign « and ». The expansion of the inverted commas will occur only if the type has been specified in the tag .
N.B.: if within a sentence with inverted commas, quotes are used once again, then they will appear as < and >. Corrections are made directly on the XML and for single angle quote the encoding characters used will be ‹ (‹) e › (›). In the control phase of the Testo Costituito the tags to be deleted won’t be signalled.

3.10 Del

Tag and content should NOT be displayed.
Anything not specified above should not be displayed on the Semantic Edition

3.11 Spaces

Spaces wider than a single unit must be encoded, otherwise in the conversion to HTML they disappear.
The TAG is:

3.12 Specific terms and acronyms on page bottom

Specific terms and acronyms indicating page turn (eg. verte, w. S. u., w. S. b. u., T. S. V. P. etc.) won’t be visualised in the Semantic Edition.

3.13 Square parenthesis together with Printed in dates

In the Semantic Edition square parenthesis are removed (only hi rend printed is kept)

3.14 Gaps and illegible characters conveyed with the symbol “%”

3.14.1 Illegible letters and words (paleographic problems, damaged paper, etc.) appear in the Semantic Edition as [-]
Illegible characters within a or are conveyed with the sign %.
One single % character is converted into [-]
A sequence of % (from two onwards) is converted into [–]

3.14.2 while the transcription character ∩ – that indicates a space between letters (due to the flow of the handwriting) – is maintained
During the transformation phase::

• the tag is converted into [-]
• the tag with quantity >1 is converted into [–]
• the symbol [-] present in the XML is kept as such

4. Conversion Chart

5. Semantic Edition checking procedure

1. XML Status validated – no odt: proceed with letter check
2. Check possible comments on the letter in the shared document
3. Open the letter on Pundit and verify comments
4. If corrections on the XML are needed: download of the latest version of the transcription, addition of the different changes and upload of the new version
   1. Changes to the letter must be recorded in order to discuss them with the group during next morning meetings
5. If needed, xslt stylesheet for convertion must undergo corrections
6. If needed, rules addition to the style sheet css for the visualisation
7. Freezing of the XML version. The field  n7tra_xml_content is copied in testo_costituito_xml: php symfony burckhardt:create-testo-costituito [id] –xml-copy
8. Testo Costituito regeneration starting from XML in the field testo_costituito_xml: php symfony burckhardt:create-testo-costituito [id]
9. Check if the XML conversion is correct
10. Check if the reported errors have been solved
11. Test (selenium or symfony) to verify the presence of the Semantic Edition

Last edited 2015-09-03