Project Euclid journal issues should all have records for table of contents, editorial staff, indices, etc., whenever the data is available. This document outlines the proper .xml encoding for these various types of records. Occasionally you will receive files of "miscellaneous" frontmatter or backmatter, where the nature of the content is such that it makes sense to create a more general record instead of the more specific table of contents or editorial staff records. Some issues may contain an index, either for the current volume, or for several volumes of the journal.
Miscellaneous Frontmatter
Table of Contents
Editorial Staff
Editorial Statements, Comments, etc.
Prefaces and Dedications
Obituaries
Articles in Foreign Languages
Related Items
Items Appended to an Article
Corrections, Errata, Addenda
Parts
Filling Out the Related Items Tracker Spreadsheet
Reviews
Indices
Miscellaneous Backmatter
a. Miscellaneous Frontmatter
Below is an example of an issue for which the publisher sent a file called "front," containing the various beginning pages besides the table of contents. Note that record type="frontmatter," and note the generic name plus publication information included in the title element. Often for frontmatter, only a .pdf is sent. This particular issue came with .pdf's and .tiff's for each item. Page numbers are seldom used for frontmatter records, as often the physical pages in the actual journal are unnumbered.
b. Table of Contents
For many issues, the only kind of front- or backmatter received is the table of contents, and for most of these records, encoding like that of the record below is sufficient.
c. Editorial Staff
The second most common form of front- or backmatter, after a table of contents, is the editorial staff page, which lists the editors for the particular issue in hand. A basic record such as the one below is usually all that is required.
d. Editorial Statements, Comments, etc.
There may be a "letter from the Editor" or an "Editorial Statement" included in either the front- or backmatter. If the editor's name appears as "author" you may include an <author> element to display this information. If not, just omit the <author> element and proceed as with other front- and backmatter as regards the formation of the title, etc.
e. Prefaces and Dedications
Follow the same guidelines as for Editorial Statements, above. The .pdfs of both prefaces and dedications should be examined thoroughly; their contents may give important clues as to the nature of the issue in hand. The examples given here are from the same issue of Methods and Applications of Analysis (Euclid abbreviation MAA). The encoding for these objects is usually quite straightforward. In this particular example, however, the preface reveals that the issue is the first of what is to be a series of issues dedicated to a particular mathematician, making it a special issue. It was determined that a special title (add link here) would be given to the issue, based on this information.
The .pdf for the Dedication:
The .xml record for this object:
The first page of the .pdf for the Preface:
The .xml record for the Preface:
f. Obituaries
An obituary may not have the word "Obituary" in its title, but it should be included in the title in the .xml record for the object. Often an obituary starts with a photograph of the deceased, followed by their name and dates of birth and death. Usually this information stands alone on the first page of the object, and the second page contains a short biography. Often a bibliography of the deceased's published works will be included as well. (See below)
The encoding is fairly self-explanatory. (This particular example also shows encoding for multiple single-page tiffs for the object, a rare practice.)
g. Articles in Foreign Languages
This section outlines the protocol for coding an article that is in a language other than English. English is the default language, so if the article is written in a language other than English, you MUST include a language attribute with the appropriate language code.
i. To cover the entire article, for instance if it's in French, code <record lang= "FR">. You should also code the title <title lang="FR" …
If the title appears also in English, include both the English and French titles, with English first, and specify the language, as <title lang="EN" …
Example: Record and title fields:
ii. Specify the language in the Abstract field also, as <abstract lang="FR" … and if the abstract is given in both the foreign language AND English, include both abstracts and put the English one first, specifying lang="EN".
Example: Abstract in 2 languages:
iii. If keywords are given in the foreign language, include them in the record as is, giving the proper language attribute. If translating keywords into English is relatively trivial, add the English ones too, with no language attribute, since the default is lang="EN". If doing the translation would be difficult, include only the foreign language keywords.
Example: Keywords:
In some issues, you will encounter articles with "related items" - commentary, etc., which are appended to the article and are not necessarily written by the article's author(s). In other cases, the "related item" is written by the author or publisher, and is published in another issue of the journal, as in a correction, erratum, or addendum. Another use for "related items" is the case in which an article is published in parts, in multiple issues of a journal.
In addition to constructing .xml records as detailed below, you will need to record most instances of related items in the spreadsheet, "Library21/Euclid_DPUBS/RelatedItemsTracker.xls". The spreadsheet was created to help document the process, to ensure that both the original article and any related articles' .xml records contain fully formed instances.
i. Items Appended to an Article
Here is the Table of Contents for an issue containing an article with "related items." Note that there is a list of articles indented under the title of the first article in the Table of Contents. These additional articles are included in the record for the first article, and are labeled as "related items" in the XML.
The XML instance for each one of these "related items" would look like the example below. The whole string of these instances would be placed between the closing of the "abstract" element and the opening of the "keywords" element.
This is how the XML created for these related items will look in the Euclid abstract display for the article:
For this example it is not necessary to make use of the "Related Items Tracker," because the related works are all published "under" the title of the leading article, within the same issue of the journal.
ii. Corrections, Errata, Addenda
When a correction is made to an article which was published in an earlier issue of the journal, you will have to create a "backward" link from the correction to the original article, and you will have to create a "forward" link from the original article to the correction. This will have to be done in several steps, as the Euclid publication number used for the link is created as the new issue is posted to the server. You can look up the Euclid publication number for the original article and include it in the <related_item> instance in the correction record, but you will have to wait until the issue with the correction record has been published in Euclid, look up the number, and then edit the record for the original article.
Seeing "Corrigendum," "Addendum," "Correction," or "Errata" at the beginning of a title in a table of contents is your clue that you will need to do a <related_item> with forward and backward links. Here is the table of contents for the issue publishing the correction in this example:
Note the form of the last title on this table of contents. Neither the word "correction," "addendum," nor "erratum" are used, but the less common term "Corrigendum" appears, followed by a colon and a title enclosed in quotes. Further inspection of the .pdf for this article revealed that it was indeed a correction for a previously published article.
To properly denote the relationship between this corrigendum and the original article, the .xml record for the issue containing the corrigendum should look like this:
The <add_desc_data> element contains the publication information for the the original article. The <identifier type="pe"> content was found by looking up the record for vol. 159, no. 1 on the Euclid website, clicking on "view abstract" for the article in question, and scrolling to the bottom of the page. The Euclid Identifier number for the article can be found as below:
After submitting the issue containing the correction, you should enter the information about the relationship to the original article in the spreadsheet, "Library21/metadata/utilities/Euclid related/RelatedItemsTracker.xls". You can mark in the spreadsheet that the "backward" link has been done. Once the issue containing the correction has been published on the Euclid site, it will also have a Euclid ID number assigned to it. You may go to the Euclid abstract page for the correction, obtain its Euclid ID number, then edit the record for the original article to include a <related_item> instance for the correction. That record would look like this:
You will sometimes find articles which are published in several parts, spread out over different issues of a journal. You may not know that this is the case with the first article of the series until the second one is published. (The clue will come in the form of a Roman numeral following the title of the article. There have been cases in which the first article's title did not contain the Roman numeral I.) You will need to follow the same protocol as in Example #2 of waiting for the second article to be published in Euclid in order to enter the Euclid ID in the <related_item> "identifier" element.
Below is an example of an article which was published in 4 parts. Notice that in this case, parts I and II were published in the same issue of the journal, and parts III and IV were both published in the next issue of the journal. You must include a <related_item> element for each part, in the record for every other part. Thus, the record for part I contains references to parts II-IV, the record for part II contains references to parts I, III and IV, and so on.
iv. Filling Out the Related Items Tracker Spreadsheet
Below are some sample entries from the "Related Items Tracker" spreadsheet:
i. Reviews Some Euclid journals publish reviews of monographic works that will never be published in Project Euclid. These works will not have links placed in their records, but will have their publication information included within the <reviewed_item> element.
Example: Here is the way a review may appear in the Euclid display.
Below is the .xml encoding for this review article. Notice that there is no cross reference, no link, to the work being reviewed. This is because this work is a monograph, and is not published in Euclid. What is included is full bibliographic information, so that the reader may find the book on their own, if they wish.
j. Indices
Author or volume indices are used in some Euclid titles, with the frequency varying from title to title. Note that record type="index," and that the title element contains information regarding the coverage of the index. This particular issue has both .pdf's and .tiff's (a separate one for each page) submitted as part of its package. The index was paginated in the hard copy of the journal, thus start and end pages are indicated in the .xml record.
k. Miscellaneous Backmatter
As some issues have miscellaneous frontmatter, others have miscellaneous backmatter. The same basic rules apply, with the exception that record type becomes "backmatter."
