• Add the argument to accept "Accession" as the program's input option. In this way, If there is no "Accession" passed in, the program will read refile barcode file. If there is "Accession" passed in, the program will read accession barcode file.
• When processing accession barcodes, the program only resets the item status from "In Transit Discharged" to "Not Charged" by using APIs.
• All of outputs are not emailed out to staff, only are saved in the log file on the server.

• Add all system jar files into NetBeans Libraries in the Properties by right click.
• After build the application, it generates the application jar file and a directory called lib that includes all system jar files.
• Copy the application jar file as well as lib directory together under the same directory on the server.

• One cataloging team's webpage has some wiggy display. They asked for help.
• I figured out that in their html file, they used two keywords in the comment line. This caused wiggy happened.
• <!--[end if]--> "end if" is the keyword phrase. Another one is <!--[if !supportLists]--> "if" follwoed by the space is the keyword/phrase. Both can't be used in the comment line. They must be changed to something else, e.g. endf, iff.

• Cataloger reported that two images in the CatStat couldn't be displayed.
• I found out there are some back-slashes (\) that should be forward-slahses(/). This made image problem happened.

• Check Voyager tables on Voyager 7 test server.
• Coordinate PST to prepare the test data.
• Prepare local batch and web application list that need to be tested, and coordinate other testers.
• Test local applications on the Voyager 7 test server.

• Update the program for another two item statuses that don't need to apply API DISCHARGE. Make multiple item status display in one field, instead of multiple rows.
• Work with other colleage to compare the result between new and existing one.
• Analyze the old one's relationship with accession barcode file.

• Create the local applications list both on web and standlone that need to be tested for Voy7 release. Coordinate the group and participate to test the applications.
• Guide other colleage to prepare testing data.
• Verify if all tables are included in the new release.

• CERL reports that the last resumption token that gets all identifiers doesn't work. But it works at Yale sites. Send the last resumption token XML file which was validate and well formed to CERL.
• The application server is shared with other 255 servers in the Yale on a fire wall. This will impact the performance speed. The very short temp database closed might be due to network problem.

• Design the application structure to be best having fastest response time when getting Voyager bib data.
• Queries are saved in external XML file. In this way, the queries can be easily read and changed.
• Tried three ways: Java Servlet with DOM, Java Servlet with out print, JSP with Java Servlet Init applied. The way of DOM is slow, because it needs to wait the whole XML file to be generated, then displays on the browser. The second way can display the results right away for the existing results, and at mean time, it continues to retrieve the data from database to add in display. However the diacritics can't display correctly. The third way uses Java Servlet's init to retrieve qualified bibids that may take a longer time. The init only runs once when starting the application in Tomcat. JSP retrieves each RSS element in the queries that have index fields. In this way, the response time is instantly.
• Solve the diacritics display problem. Diacritics can't display correctly in Java Servelet. They can display correctly in JSP. The method should use: new String(rsTitle.getString("rss_title").getBytes("8859_1"),"UTF-8").

• There is holding data in SFX database, but some of them can not be displayed out. Figure out that some fields have the strange character. This character doesn't show anything, but the length is not 0. So change the program to judge the fields length is 0 or not.

• Study two books to know and understand the functions.

• Attend introduction seminar.
• Read brochure to understand its content and procedure.

• Use an example to explain. DS1, DS12, DS2, DS112, DS26, DS3.
• The computer will sort them as DS1, DS2, DS3, DS12, DS26, DS112. But this order is wrong in LC sorting rule.
• The correct sorting should be DS1, DS112, DS12, DS2, DS26, DS3.
• In order to achieve this by computer, it need to add spaces(use b to present a space) between letters and digits to make the lenght is 7.
• DS1 shoud be DSbbbb1. DS112 should be DSbb112. DS12 should be DSbbb12. DS2 shoud be DSbbbb2. DS26 shoud be DSbbb26. DS3 shoud be DSbbbb3.

• Since Eudora is replaced by outlook. The incoming mail server's domain name is changed. Central campus's was used to be netid.mail.yale.edu, and medical campus was email.med.yale.edu. Now both are connect.yale.edu. This change caused the problem that the email couldn't be sent out.

• Display the reserve charges for the selected semester if the selected semester is after Fall 07 (including Fall 07). Otherwise display the current reserve charges for the purpose of reference.
• "Search By" result only shows reserve list title with check box, "SELECT ALL" box, "Preview", "Export" buttons. Detail data that include bib, holding and item data will be retrieved by clicking "Preview" or "Export" button. Detail data only is only extracted when the list title is checked.
• Set up relations between the content of "search by" and term code. e.g. at MUSIC Course Reserve Desk, there is only term code "SP04" pop up. That means at this location only includes "SP04" reserve list, no other semester has reserve list.
• Searching input area should be "search by" box appeared before term code input box.
• Only search locations that have reserve list, not all locations in the libraries.
• Clear out the display area on the right when clicking the menu on the left.
• Highlight the menu selection after clicking.
• The detail data display The link function effects Ajax functions that have been used. Must consider clicking behavior.

• Remove summer term code from Undergrad Senior and Grad Student. TERMCODE format still is YYYYSS (200801).
• Undergrad senior:
  SS= 01 (From Jan.1 to Aug.15 for spring)
  SS= 03 (From Aug.16 to Dec.31 for fall)
  
• Grad student:
  SS= 01 (From Jan.1 to Jun.30 for spring)
  SS= 03 (From Jul.1 to Dec.31 for fall)
  
• Other patron groups:
  SS= 01 (From Jan.1 to May 31 for spring)
  SS= 02 (From Jun.1 to Aug.15 for summer)
  SS= 03 (From Aug.16 to Dec.31 for fall)
  

• Figure out that incorrect PATEO with six segments causes the problem.
• Work with PST to query out all incorrect PATEOs from fund table for LBO to correct.

• In the html, apply DIV for the layout; CSS for the appearance; Ajax Prototype and Scriptaculous Autocompleter function.
• In the JavaScript:
  1. Random number in the URL to avoid browser's caching.
  2. "&" is used as delimiter of URL parameters, it can't pass over as content between client and server, it must be replaced with its encoding by using replace(/&/g,"%26") when passing into server or replace(/%26/g,"&") when getting data from server.
  3. Make connection between client and server by using myRequest.getXMLHttpRequest(), myRequest.responseXML; Control click action by using Event.observe.
  4. Use document.getElementById function to pass values between html and Javascript.
  5. Use xmlResponse.getElementsByTagName to get data from server.
  6. Write html syntax into the string variable, then write back to browser by using document.getElementById('ResultDisplay').innerHTML= '<div id="ReturnResponseContainer">' + variable.
• Some JavaScript can not be used in both IE and Firefox.
• The JSP that relates with Autocompleter retrieves data from Oracle database by using LIKE search. Any typing will invoke query searching database by the condition LIKE '?%'
• The JSP that relates with searching reserve list data has these features:
  1. In the Oracle query "&" is treated as parameter input. If the input includes "&", it must be replaced by "&'||'" for the query.
  2. Use HashMap, and Hashtable save data from department, instructor, course, reserve location Oracle tables. So that they don't need to directly participate in the query to make searching much faster. The "&" that comes from Oracle table is saved as "&" in the HashMap and Hashtable. It needs to be converted back by using function replaceAll("&", "&").
  3. Queries are saved in external XML file. The input part is replaced by each searching category in the JSP. In the query, use right join (+) to avoid missing the records that don't have data in some fields.
  4. Two searching modes: "StartsWith" whose query uses "LIKE '?%'" condition; "Exact" whose query uses "=?" condition.
  5. Term Code is from reserve list title's data between first and second space.
  6. inputData.replaceAll("&'||'", "&") doesn't work correctly. It should use inputData.substring(0,inputData.indexOf("&'||'")) + "%26" + inputData.substring(inputData.indexOf("&'||'")+5,inputData.length());
  7. Syntaxes of return data back to browser:
     response.setContentType("text/xml");
     out.println("<?xml version=\"1.0\" encoding=\"UTF-8\"?>");
     ......
  8. Use session to pass query results, HashMap, Hashtable for the JSP that exports data into MS Excel sheet.
• The JSP that relates with export to MS Excel sheet receives the data from session, and:
  1. html codes don't be recognized in the Java's "if" condition <% if {...} %>
  2. Use jstl core library to loop the result set, and convert data between jstl, Java, html, such as: <c:forEach var="row" items="${getRecordQuery.rows}" begin="0">
     <c:if test="${reserveID != row.reserve_list_id}">
     <c:set var="dept" value="${row.department_id}" />
     <jsp:useBean id="dept" type="java.math.BigDecimal" />
     <% out.println(((dept.toString().equals("0")) || deptHash.get(dept) == null) ? "" : deptHash.get(dept)); %>
• Use CSS to control colors, typography, layout schemes, translate into greater power and efficiency, create cleaner, standards-compliant code, improve the accessibility and usability.

• Interface layout, color, font, header, footer, menu, display area, convenience and easily use.
• Interface friendly input design will use Ajax Prototype and Scriptaculous autocomplete to prompt what the database has. This will greatly help user for their right input. Use CSS to create nice readable interface.
• Client side Javascript connects interface's activities into server side actions. Javascript gets interface's parameters'value; processes to generate new values; passes them into server; gets response back from server; if server returns the result, parses, and write into interface to display, otherwise write waiting statement.
• Server side JSP gets request from Javascript, search data from Oracle database, wrap the results into XML to return into Javascript. In JSP, it uses Java technique to make less Oracle tables involved into the runtime querying to make searching time as fast as possible .
• Analyze, test Oracle tables to create the efficient queries.
• Easy export to MS Excel format feature.

• Go over the request with System Librarian to understand the need.
• Research Voyager Oracle tables to find the relationships among tables, fields, contents that relate with course reserve list.

• Organize the Jira review meeting for all library Jira users.
• Give the Jira demo to go over Jira's features, in order to help colleagues understand Jira better.
• Encourage colleagues to share Jira knowledge to others, so that more Jira features can be discovered, and used.

• Go over each example to test and learn its function in order to understand the technique that is applied on the example.
• Read books and class materials to learn how to use Ajax Prototype and Scriptaculous.
• Test autocomplete, responseXML, responseText in Javascript use, click observe, DOM parsing and adding.

• Analyze the programs and Voyager Oracle tables and Voyager ACQ client to figure out the problem.
• Find out that the problem wase caused by ACQ dept staff not input discount price into the correct field.

• Provide technical support to help MSSA verify their collection identified by value of encoding level.
• Update the SQL query to recognize the MSSA collection.

• Not to use API DISCHARGE items whose item status is "On Hold".
• Add one column in Excel exception file "Item Status After Discharged".

• Develop Ajax-powered interactive and dynamic Web pages.
• Design accessible interfaces for cross-browser compatibility.
• Leverage toolkits to rapidly create rich user-friendly interfaces.
• Optimize and strengthen code to build stable applications.
• Protect vital information from interception.

• Write Self-evaluation for March 2007 - March 2008.
• Write Work Journals for September 2007 - March 2008.

• Becuase Tomcat 5.5 doesn't set default to make display directory list on available by using browser, web application DownloadFile can't work correctly. The users can't have file list for them to download their files.
• Search web to find information and give the recommendation for ITS-LSG to consider.
• Work with the group trying to find the different options to solve the directory listing problem. The options are turn on the directory list in web.xml; Set up the directory list being visible in browser in apache httpd; Develop local web application to display the directory list.

• Find out that another Voyager API CIRCDEFS must be applied for discharge location before DISCHARGE API.
• If CIRCDEFS is applied a circ desk location, DISCHARGE's status is "discharged"; otherwise is "Not Charged".
• Overdue items can be DISCHARGEd, and the amount of fine/fee will be in the result of DISCHARGE.
• If the item has other item status, such as hold, recall, lost etc will not be DISCHARGEd.
• Work with colleagues to review the test results.

• Find out following mistakes that didn't obey ACQ guideline.
  • Wrong: Invoice's comment didn't write at the first line in the note field.
  • Correct: Invoice's comment should write at the first line in the note field.
  • Wrong: Multiple funds paid the multiple copies for one line item.
  • Correct: If one invoice line item paid by multiple funds needs to have multiple copies, each copy has to be made as its own independent line item with a single copy included.

• Confirm all web applications are ready to be promoted.
• Organize the promotion schedule
• Trouble shooting the problem that happend during promotion period.

• The bound-with book only has one barcode to map multiple bibliographic records. Some of Kirtas items are bound-with books. Kirtas' OAI identifier is item barcode. One OAI identifier only can match one bibliographic record. I report it to the Mass Dig team to decide if the bound-with books should be selected for Mass Dig collection.

• Find the solution for Tomcat web application to reconnect datasource automatically without re-starting Tocmat. The configuration of Oracle and MySQL is different.
  • For Oracle, there two situation:
    1. If the datasource is configured in server.xml(context.xml), add testOnBorrow="true" under "Resource".
    2. If the datasource is configured in Java program, use setAutoCommit(true).
       con = DriverManager.getConnection("jdbc:oracle:thin:@magellan.library.yale.edu:PORT:SID,database_name,database_password);
       con.setAutoCommit(true);
  • For MySQL, add "?autoReconnect=true" at the end of datasource URL in web.xml.
    <init-param>
    <param-name>dbURL</param-name>
    <param-value>jdbc:mysql://exlib2.library.yale.edu:3309/sfxtst3?autoReconnect=true</param-value>
    </init-param>
    
• Find the solution to make two datasources on two servers alternate work if one server is down. Put primary datasource in Java "try" section, and put secondary datasource in this "try"'s catch block. The primary datasource will be connected all the time unless it lose the connection; then the datasource connection will be automatically switched to secondary datasource. Whenever the primary datasource is back to work again, the datasource connection will be automatically switched back to primary datasource immediately. This approach can make applications work 24 hours 7 days if two servers are not down at the same period.

• Work with ITS-LSG for unix script that is used to run the program, and read LSF manifest files to count Kirtas item.
• Update the program based on test results.

• Try to find out why the project select box is missing.
• Test JIRA feed function with co-worker's help, try to find the feed problem.
• Try to find out how to set up public users.
• Try to find out how to create the issue that open to public.
• Find out the group user permission of issue type configuration.
• Contact JIRA technical support for the necessary help.

• Schedule and make sure all applications are in their place to be ready roll over to the production.

• Create Oracle query.
• Develop codes adding in Circ Reports as a new function.
  Select item type code through drop down list.
  Display in the browser, or export to the Excel sheet.
  

• Create local table to hold candidates of Kirtas item barcodes.
• Develop the program to import candidates of Kirtas item barcodes.
• Develop the program to extract the given number of barcodes that don't have pull out date; update the pulling date as current date; generate LSF barcode file that formates as GFA needs; save into GFA specific directories.

• Change old OAI metadata prefix oai_marc to the new one marc21.
• Change the data format from oai_marc to marc21.
  oai_marc format sample:
  <oai_marc status="n" type="a" level="m" encLvl="7" catForm="a" xmlns="http://www.openarchives.org/OAI/1.1/oai_marc" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://www.openarchives.org/OAI/1.1/oai_marc http://www.openarchives.org/OAI/1.1/oai_marc.xsd">
  <fixfield id="1">" 90178038 "</fixfield>
  <fixfield id="8">"900724s1974 po af 000 0 fre "</fixfield>
  <varfield id="40" i1=" " i2=" ">
  <subfield label="a">DLC</subfield>
  <subfield label="c">DLC</subfield>
  </varfield>
  <varfield id="300" i1=" " i2=" ">
  <subfield label="a">168 p., 67 leaves of plates :</subfield>
  <subfield label="b">ill. ;</subfield>
  <subfield label="c">33 cm.</subfield>
  </varfield>
  </oai_marc>

  marc21 format sample:
  <marc:record xmlns:marc="http://www.loc.gov/MARC21/slim" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://www.loc.gov/MARC21/slim http://www.loc.gov/standards/marcxml/schema/MARC21slim.xsd" type="Bibliographic">
  <marc:leader>00590nam a22002051 4500</marc:leader>
  <marc:controlfield tag="001">2445743</marc:controlfield>
  <marc:controlfield tag="008">010411s1805 enk 000 0 eng d</marc:controlfield>
  <marc:datafield tag="040" ind1=" " ind2=" ">
  <marc:subfield code="a">YUS</marc:subfield>
  <marc:subfield code="c">YUS</marc:subfield>
  </marc:datafield>
  <marc:datafield tag="300" ind1=" " ind2=" ">
  <marc:subfield code="a">2 p. l., [3]-94 p.</marc:subfield>
  <marc:subfield code="c">24 cm.</marc:subfield>
  </marc:datafield>
  </marc:record>

• Attend this conference, here is my note.
• JMaki
  • JMaki wraps many existing cool Ajax tools together, such as JSF, Dojo, GWT(Google Web Tools) Yahoo ajax tool, FishEys, TreeNodes etc in NetBeans 6.0. You just drag them into coding area, the NB will generate codes structure for you. Then you can add yourself codes.
• Profiler
  • NB 6.0 builds in Profiler which can monitor threads, CPU, memory leaking problem, and can take you right to where the bad code is, and other more.
• Java FX
  • A new scrip language which is just announced at this year's JavaOne conference(Jeff and Roy went this one, right?). Mainly for web design. The concept is: When the web designer is not the web developer, it's very hard for the web designer to write Java code. With Java FX, the web designer doesn't need to write pure Java code, Java FX will do for you.
• Ruby
  • In NB, it's called JRuby(Java Ruby). JRuby is built in NB 6.0.
• Java DB
  • It called Java Deby database. It is Sun's open source. Java DB is built in NB 6.0. Maybe in the future, we can build our local tables in Java DB, instead of in Oracle.
• XSLT project module, and XSLT editor that looks like XMLSpy style.
• SQL project module, and pull out the results with XML format option.
• Kill the running job at output window. This is very helpful. Right now, I have to close entire NB in order to kill the job.

• Figure out that maximum number of line items exceeds the original number. This caused array out of range.
• Figure out the wrong PTAEO number (One more segment) was entered inadvertently from Voyager client. This also caused array out of range.
• Add sslegc into SSL, instead of SML.

• Build up JIRA production settings. Create pioneer projects and issues.
• Figure out the way to make projects as inactive archive status.
• Create a special project as the home of all documenations.

• Read the guides of XMetaL Author and XMetaL Developer, and watch their demoes.

• Standard JIRA has Notification and Permission two schemes. Enterprise JIRA has more schemes.
  1. Notification Schemes are used for sending email notifications. Different events can be added into one notification scheme that make what activity should send email.
  2. Permission Schemes are used to define user's privileges in the project. This relies on the configurations of users, user groups, and project roles.
  3. Issue Security Schemes are used to control who can and cannot view issues. They consist of a number of security levels which can have users/groups assigned to them. When creating the new project, add Issue Security Scheme and Security Levels on this project. In this way, the issues can only be accessed by the users/groups that are assigned in security levels.
• JIRA doesn't have real Archive feature. By adding a permission scheme that doesn't link to any users, groups and roles, it can achieve the "Archive" feature. Whenever the project is done, change the project's permission scheme to this "Archive" scheme. Then this project won't be showed in any users' login project's list. It's only displayed in Administration project's list with the red colored in "Default Assignee" field. If this project needs to work again, change the "Archive" scheme to other working permission scheme.
• JIRA has three level's hierarchy structure: project -> issue -> subtask.
• JIRA can plug in Subversion. atlassian-subversion-plugin.jar*, javasvn-.jar*, ganymed-.jar* need to be placed in WEB-INF/lib directory, and local Subversion server data needs to be put in a properties file subversion-jira-plugin.properties that should be placed into WEB-INF/classes directory.
• JIRA has two ways to create public user account. This is controled by JIRA two modes which are in General Configuration:
  1. Public - Any user can signup and post issues.
  2. Private - Only administrators can create new users.
  By combining the configuration of security levels in Issue Security Schemes, the public users can be restricted to access issues that they create only. The public users can see all project's names. When they click each project, they can't see any issues that are created by others. However, the public users can create the new issues on all projects.

• Solve the problem in NetBeans5.5.1 + Tomcat 5.5 that JDBC driver can't be found.
  1. Place ojdbc14.jar into Program Files -> netbeans-5.5.1 -> enterprise3 -> apache-tomcat-5.5.17 -> common -> lib.
  2. Change context.xml to tomcat5.5 format:
     <?xml version="1.0" encoding="UTF-8"?>
     <Context path="/CurrencyJSP">
       <Resource
           auth="Container"
           description="DB Connection"
           name="jdbc/xxxxxx"
           type="javax.sql.DataSource"
           driverClassName="oracle.jdbc.OracleDriver"
           url="jdbc:oracle:thin:@xxxxxx:1521:xxxxxx"
           username="xxxxxx"
           password="xxxxxx"
           maxIdle="2"
           maxWait="5000"
           maxActive="4"/>
     </Context>
     
• Plug in Subversion in NetBeans5.5.1:
  1. Download SVN client from http://subversion.tigris.org/servlets/ProjectDocumentList?folderID=91
  2. Go to NetBeans Tools -> Update Center -> check NetBeans Update Center, Next -> Select Subversion -> Add, then install.
  3. Subversion tab will appear next to CVS tab.
  4. Click Subversion -> Checkout -> in Repository URL, enter http://xxxxxx/svn/repository/
  5. Enter username and password, find the project that you want to check out by "Browse", select Local Folder, then Finish.
• Plug in Maven in NetBeans5.5.1:
  1. Download Maven client latest version (Maven2 Support) from http://mevenide.codehaus.org/m2-site/index.html.
  2. Unzip .nbm files into one folder.
  3. Go to NetBeans Tools -> Update Center -> check Install Manually Downloaded Modules (.nbm Files), Next -> Add all unziped .nbm files, then install.
  4. Installation reference URL: http://mevenide.codehaus.org/m2-site/mevenide2-netbeans/installation.html
  5. Build maven web application sample URL: http://mevenide.codehaus.org/m2-site/mevenide2-netbeans/tutorials/helloWorldWebApp.html

• Learn Ruby, Rails, Ajax, and Pro PostgreSQL, and NetBeans 5.5.1 plug in Subversion and Maven.

• Research OAI MARCXML standard protocol.
• Describe Lewis Walpole Orbis OAI Extract brief technical procedures.
• Describe what work need to be done, potential application that will use this tool, and time frame.

• The single quote in MySQL search symbol is \'. But in the Java programming, \ must be represented by \\. So the single quote in programming must be represented as \\\' in order to be recognized.
• Make catch block as normal XML node display.

• Create seven annual circulation Oracle queries. This is an enhancement from past MS Access queries. The speed improvement from two hours in MS Access down to current less than 5 minutes for each of queries.
• Add coding into Circ Reports application package for these seven annual reports.
• Use Oracle 9 full outer join to join two different tables with the same field names, but two tables are not have any link relationships. Full outer join joins all records including the duplicate records. Union join only joins the different records. If the records are identical, only one identical record are joined.
• Trouble shooting Final Search report. The problem is that temp loc id in item table is 0 which is not defined in location table. When query was trying to get the location name, only find those whose temp loc id is not 0. Because most of temp loc id in item table is 0, many records were missed. Update the query as display all no matter location id is defined in location table or not.

• Add GOV DOCS into SSL, instead of SML.
• Add condition to separate the staff from central campus and medical campus, because two campus's email incoming domain names are different. In this way, staff from all librares can receive invoice transfer snapshot excel files that send by invoice transfer program's sending email function.

• Learn SFX, and MySQL.
• Download and install MySQL Query Builder.
• Analyze SFX database tables, and the relationship among the tables.
• Create keyword search, category/subcategory search queries, and document the analyzing thoughts for other staff to share.
• Use Java Servlet to program the application.
• Specially use Java Regular Expressions to code translation from SFX holding availability raw data to readable data.
• Code to merge SFX holding availability local and global data.
• Figure out Java servlet code to output UTF8 data correctly in XML format.
  String CONTENT_TYPE = "text/xml; charset=UTF-8";
  response.setContentType(CONTENT_TYPE);
  DocumentBuilderFactory factory = DocumentBuilderFactory.newInstance();
  DocumentBuilder parser = factory.newDocumentBuilder();
  Document doc = parser.newDocument();
  OutputFormat format = new OutputFormat(doc, "UTF-8", true);
  
• Document "How to use this application - URL explanation".

• For the months since November 2006, downloaded and installed different SQL Developer versions, due to bugs in the SQL Developers. Test using its functions, especially test "Query Builder" function. Find the problems, and contact with Oralce expert for the solutions. After Oracle issued a few new patch releases to fix the problems, now SQL Developer 1.1.3 is ready for us to use.

• The invoice transfer file only includes the note that any texts following "CMT=", and in the note’s first line (ended by ‘\n’), exclude ‘;’ will be transcribed in the transfer file’s Invoice Description (or Note field). But one invoice note didn't obey this guideline, it puted "CMT=" line on the second line. That caused the invoice transfer job ran failed.

• Get major bib and holding non supressed records of BRBL's all locations.

• Voyager's Oracle database has US7ASCII defined as NLS_CHARACTERSET. This affects all VARCHAR2 and CHAR columns. Tables which contain MARC data, like bib_data and bib_text, store that data in these US7ASCII VARCHAR2 columns. All MARC data is encoded as UTF-8, a sequence of 8-bit bytes. Oracle does store all 8 bits of each byte, so the UTF-8 encoded MARC data is *stored* safely in these US7ASCII columns. It's just that when most programs ask Oracle for this data, Oracle reasonably enough returns the data as US7ASCII since that's what the NLS_CHARACTERSET says is correct, which causes the diacritic data to be corrupted.
• In Voyager, only MARC data is encoded as UTF-8. Other data, like patron names, can have diacritics, but all the non-MARC data is encoded as ISO8859-1.

• Annual Circ Transactions by Item Location.
• Annual Circ Transactions by Item Type.
• Annual Circ Transactions by Patron Group.
• Greatly improve the query's efficiency compared with MS Access queries that ran in previous years, from 2 hours running time before to current less than 5 minutes.
• Find two problems:
  1. By the date annual circ reports running, some of item records in circ transaction tables were deleted from item tables.
  2. One patron group in circ transaction tables is not defined in patron group table.

• Re-query a few circ reprots' queries based on circ dept. requests.
• Add a new tab for LSF reports.
• Add new function on a few circ reports.
• Help solve the problem when opening Excel file in IE:
  Tools -> Internet Options -> Security -> Custom Level -> Downloads -> File download -> Disable
• The code in JSP to display UTF-8 in browser:
  <HEAD>
     <%@ page contentType="text/html; charset=UTF-8" pageEncoding="UTF-8" %>
     <META http-equiv="Content-Type" content="text/html; charset=UTF-8">
  </HEAD>
• The code in JSP to open Excel file in UTF-8 :
  <HEAD>
     <%@ page contentType="application/vnd.ms-excel; charset=iso-8859-1" %>
     <META http-equiv="Content-Type" content="text/html; charset=UTF-8">
  </HEAD>

• Consult with Oracle expert about sqlDeveloper language encoding feature, and experiment our need using this feature.
• Consult with UCLA's expert about Voyager encoding envoironment.
• Discuss with ITS about above two issues.
• Contact with other institutes and university about qualified OAI DC open source.

• Communicate with related staff to get the requests.
• Query the function.
• Coordinate the procedure of running query and transfer results.

• Use SQL PrepareStatement to run the query instead of using CreateStatement. In this way if the query in the loop, only loading the query once.
• Use jspInit() for the one time run job. In this way, the one time job only is executed once at the first time loading the JSP page; then stay in the cache. This method can improve the performance.
• Update the LC class range design as two level hierarchy structure with category id to group the categories, and avoid to oupput the same record repeatedly.
• Use AJAX to build the connection between server and client. The client (browser) issues the request, and server runs the queries. The queries are built dynamically based on each request. The results will be written to a txt file, and saved on the server. The file name is timestamped with netid.
• The user will be notified in a email when the file is available. The user click the link in the email, then the user will be taken to the directory that her/his file is saved.
• Use Java Mail to send the email to the user automatically. There are two email domain server names on the campus. Use user profile group name to decide the user's email domain server name.
• The txt file is delimited by pipe "|". The txt can be imported to Microsoft Excel or Access.
• It is recommended that number of output records should be less than 40,000.

• Discuss with the group about the concept, function, procedure, workflow, feature.
• Provide estimation level of effort for the development.

• Create a web apps structure to allow this apps' root can be reached through URL.
• Create a directory to in which the files can be saved.
• Connect to this URL provided by email.

• Read and learn SQL Developer functions.
• Install, test use it and find problems.
• Communicate with Oracle support to report the new release bugs.

• Update the programs for the CAT new requirements.
• Coordinate and provide to CAT technical supports from different aspects, and move it to production environment.

• Provide technical support to CIRC.
• Communicate with all parties and help make the project forward.

• Summary the issue to be happened in three categories: database, browser, files.
• Reseach the possible solution, and suggestion from the related technical forum.
• Contact with Oracle gurus to answer my questions.
• Present the technical ideas to the group.
• The browser should only response the first rows for the brief display purpose as very fast as it can.
• The full complex queries should run background without the user interaction. After it is done, notify the user.
• Prevent database error by limiting user's selections.

• Include four aspects about this tool: Objective, Benefit, Overview, Description of Practical Applications by using this tool.

• Document Cataloging Statistics technical aspects.
• Finish last changes, re-jar, and re-deploy to the server.
• Coordinate for promoting it to the production project.

• Update the program to use 024I or ISB3 as the matching link.

• Read Fedora tutorial, and go over the tutorial demos.
• Share self comprehensive understanding to the group.

• Update to create two Voyager views.
• Document views' description.

• Finish the developing work.
• Test invoice transfer files, and snapshot reports.
• Write the documentation for invoice transfer project.

• Running the each project's program under the new system environment.
• Work with PST, and ITS staff to check the results.

• Analysis, and discuss the policies, standards, and request with ITS and library staff.
• Create Oracle SQL queries to reflect them.

• Analysis the request of E-Resource of Cataloging Statistics.
• PhotoShop the new tabs for each category.
• Create the new tables for each category.
• Programming the function.

• Investigate the relations among acquisition database tables that relate with invoice, fund, line items, other charges.
• Master both library and account payable's requirements and criteria for invoice transfer.
• Investigate the problems that sometimes happen in the invoice transfer files caused by Voyager's bug.
• Investigate the problems that sometimes happen in the invoice transfer files caused by foreign currency conversion.
• Use Java Excel API to create snapshot excel sheet reports. Open source jar file: poi-2.5.1-final-20040804.jar.
• Design the effient, and flexible programming methods to create invoice transfer files.
• Use Java mail function to send snapshot excel sheet reports to library staff through emails automatically.
• The open source jar files are used for emailing function: mail.jar, activation.jar.

• Review work flow, and understand all parties'requirement.
• Provide technical support and explanation to circ.

• Add the ability to retrieve the given peroid total.
• Add the calendar ability to pick up start and end date for calculating statistics total.

• Develop webform to replace the manually entered paper forms that cataloging department is currently used for statistics.
• Create 18 local database tables to storage different category statistics data. Each record is identified and unique by its netid and operation date.
• Use AJAX technique to handle add/subtract operation instantly without refreshing webpage.
• Apply CAS authentication to group staff access to different statistics categories.

• Design and develop the interface.
• Design and develop the input selection data and rules.
• Develop the dynamical queries that are create by program based on the input selections.

• Create eleven material category tables and one staff profile table.
• Create header logo and category tabs using PhotoShop.
• Develop tab switch function in Javascript.

• Create all queries for generating lists from four select boxes.
• AJAX, Java thread, hashmap, treemap, collection are used to the function of process result sets, progress monitor display.
• Create three local tables to hold LC call numbers information.
• Create header logo, title icons, buttons using PhotoShop.
• Apply CSS on the interface.
• Create footer.

• Go over the project's content and expectation.
• Study the concepts that related with this project.

• Use AJAX, Java HashMap, TreeMap, Collection to sort, select, execute call number and location data.
• Select Location(s) - sort as location code; value is location code.
• Select Class(es) - sort as location code, call number class; value is location code + class.
• Select Subclass(es) - sort as location code, call number subclass; value is location code + subclass.
• Select Range(s) - sort as location code, call number range; list display with call number range's hierarchy and sequence.

• Voyager invoice id's data type is digit and its API accompanies the digits.
• Old system invoice id's data type is character, and for some reason it comes to the payment file whose check number needs to be added into Voyager.
• Because the data types' difference, API can't process correctly. It adds the check number to the wrong invoice record.
• Add "filter" to check invoice id's data type, so the data with character type can't go through API.
• Write this id into problem file for ACQ staff to check.

• Use Java to create and process XML documents.
• Use Java to build a working prototype of a Web-based application that can generate, exchange, format and control XML.
• The technique is good for exchanging data over the Internet.

• Write server running Read Me file.
• Write programming procedure, matching criteria manual.
• Organize old version writings.

• Learn Cognos Business Intelligence Demo.
• Endeavor Analyzer is the product that applies Cognos Business Intelligence onto Voyager, currently onto Meridian.
• The relationships in presentation layer are predefined. You can only get reports that fields are in presentation layer.
• Other reports that fields are not in presentation layer may can get through database schema.
• The output formats are very flexible, such as HTML, EXCEL, PDF, DISHBORAD, PIE GRAPH, BAR GRAPH, etc.

• Voyager normalized call number uses spaces, fixed length to make call number sort correctly. The first part of call number is 7 length. It includes letters, spaces, digits. The number of spaces depend on the total number of letters and digits.
• When searching with call number range, add adjustment spaces between call number letters and digits to prepare as 7 length format. The ending call number range uses '%' at the end of range.

• Start from call number search model design.
• Use AJAX technique that when query runs on server, client still can do something else.
• Design four layers for call number search: location->LC class->LC subclass->LC subclass range.

• Match from BIB_INDEX 020N, 024A. If length is 13, check if the first 3 digits are 978 or 979.
• Use open source james.
• Take the first found ones if found multiple qualified records from both YALEDB AND LCDB.

• Design and create LC class table, LC subclass table and LC subclass range table.
• Class table has two fields - class letter and class label. Index is class letter field.
• Subclass table has three fields - class letter, subclass letter and subclass label. Index are class letter and subclass letter fields.
• Subclass range table has four fields - subclass letter, start number, end number, and subclass range label. Index are subclass letter, start number, and end number fields.
• These tables are used for new collection analysis tool.

• When query is against large Oracle table, query's structure is very important. It determined the performance speed.
• The first step should narrow down subset as smaller as possible.
• If there are multiple tables involved, the order of table appearance/joint relations are crucial too.
• When query function is applied into query, the query speed will be impacted based on the function. Some functions will tremendously slow down query's execute speed, e.g. substr.
• Oracle Database 10g supports Regular Expressions.

• Document MARC and OAI metadata match table.
• Apply the new query to generate finding aids records.

• Write programming procedure and Read Me documentations.

Resarch and Test AJAX, Mutiple Threads Combined Example

• Create an example - progress bar for waiting SQL query running on the web browser.
• Place dwr.jar in the WEB-INF/lib.
• Create a dwr.xml file that lives in WEB-INF alongside web.xml. dwr.xml makes connection to pass data between server side Java and client side JSP/HTML.
• Edit web.xml to make dwr.jar be able to be used.
• Includes links to the JavaScript files that make the magic happen in the JSP/HTML file.
• Develop the related standalone Java programs and place them in the WEB-INF/classes.
• There are two thread java programs in the example. One thread is a process of running a SQL query. It sets the different value when query is completed or not . Another thread is used to monitor the data value of if status that query is completed.
• The magic JavaScript files have two types: system JavaScript files - util.js, engine.js and application JavaScript files.
• The application JavaScript file includes following features:
  • Invoke threads by calling Java program to start threads.
  • Monitor query completed status by repeatedly calling the thead that monitors query progress.
  • Generate the web browser's progress bar and related text messages.
  • Display final query results.
• Research and make Tomcat data source that declares in context.xml available to standalone Java program by this way:
  • Context initContext = new InitialContext();
  • Context envContext = (Context)initContext.lookup("java:/comp/env");
  • DataSource ds = (DataSource)envContext.lookup("jdbc/ReportsDB");
  • c = ds.getConnection();
• Test to demonstrate that Tomcat Realm can work under this combination technique.

• A thread (sometimes called an execution context or a lightweight process) is a single sequential flow of control within a program.
• Threads are used to isolate tasks.
• Each thread is a sequential flow of control within the same program.
• Each thread runs independently from the others but at the same time.
• The life cycle of a thread is: creating a thread, starting a thread, making a runnable or not runnable thread, stopping a thread.

• Learn some important principles of Web design that apply to just about any purpose or theme. This course explores several pitfalls of Web design and how to avoid them, and introduces a variety of design tips and guidelines that you can use no matter what your Web site project entails.

• Use two APIs - GETINV, SEXPD to handle adding check number, check amount, check date in the database.
• Package, test on magellan, and commit to CVS.
• Write documentations for this project.

• Review current Collection Analysis programming structure, flow procedure, and problems.
• Read new related information to get ideas.
• Set up and listen to advisory group's opinions about current CA.

• Introduce concept of Xpath, Templates, Production, Transforms, XSLFO.

• Analyze the non-English Unicode display problems. It can be related with multiple factors from programming, browser, and Oracle.
• The sample code for JSP to handle unicode is following:
  1. <% @page contentType="text/html;charset=UTF-8" %>
  2. <% @page pageEncoding="UTF-8" %>
  3. <html>
  4. <head>
  5. <meta http-equiv="Content-Type" content="text/html;charset=UTF-8" />
  6. </head>
  7. <body>
  8. ??(Hello)
  9. </body>
  10. </html>

• Learn the concepts of acquisition to process invoice.
• Learn the procedure of invoice transfer.
• Learn the structure of invoice transfer programs.
• One thought about future updating invoice transfer program: develop web application to find transfered invoices, and replace current email mode.

• Learned how to develop and maintain portable scripts useful for system management, data manipulation, and Web CGI programming.

• Update an OAI open source to process finding aids records from Orbis database.
• The format of input records is MARC21 and the format of output records is oai_dc.
• This stage is pilot experiment and test period. It demonstrates that it's possible to harvest data from Orbis tables.
• The pilot test has been successfully harvested by MetaIndex and records are searchable through MetaLib.

• Read university invoice file to get check number, Expenditure date and other related information.
• Use Voyager API to write into Orbis database.
• One issue needs to be attention: the amount of invoice can be different from one item's price that shows up in Orbis, because multiple items can be paid in one invoice.

• Research which Voyager APIs can proceed this function.
• Test these APIs.

• Test web-based applications.
• Test standalone fund allocation application.

• Documentation for Fiscal Year Funds Excel Spread Sheet (FYFundLedgerExcelManual.doc).
• Documentation for New Fiscal Year Fund Allocation (FundNewAllocationManual.doc).

• Purpose: gain knowledges to re-design and re-develop Yale collection analysis tool.

• Purpose: convert invoice transfer from current Visual Basic to Java environment.

• Read, learn the documentations about 13 digit ISBNs.
• Voyager index 020A for matching entire character string and 020N for matching number only.
• Use 020N as matching point between Orbis and LC.
• Propose the request for updating programs - bib id list and which 020N should use if multiple 020N exist.

• Implement UIUC OAI oai.jsp open source.
• Find and fix the bug of ListRecords at from date input.
• Make this tool available in MetaIndex.
• Meta data saves in MySql. This tool supports OAI set function.
• Implement OCLC OAICAT open source.
• Test filesystem mode. In this mode, one XML file only includes one meta data record. In this way, this tool doesn't support OAI set function.

• Research, read, learn OAI documentations and OAI websites.
• Download and install OAI data provider open source tools.
• Install and set up MySql environment.
• Create program and use MARC4j to load MARC records into MySql database being used as OAI data repository.
• Read open source code to learn how OAI data provider programs work for standard verbs.
• Update open source code and configuration to fit local needs.
• Use OAI-PMH tester http://re.cs.uct.ac.za/ to harvest testing OAI data from MySql and XML repositories that are generated by customized OAI data provider testing tools.

• Read two release documentations to understand what's new.
• Update and test local critical applications under the new system.

• Update bursar transfer programs to have either SSN or UPI in SFAS AR files.
• Update replacing patron barcode prgram to locate selected patron by UPI.

• Develop and test the procedures and standards of tomcat JDBC realm.
• Document the whole implementing process and build the templet for other staff members in the group to use.

• Develop the application to manipulate tomcat JDBC special user roles table, including ADD, UPDATE, REMOVE, and RESET function.

• Cascading Style Sheets(CSS) is an important component of successful Web design. It is integrated into HTML 4.0, so to create valid HTML 4.0 pages, we should use CSS to control our colors, typography, and layout schemes, rather than the HTML style-oriented tags and attributes of previous versions of HTML.

• Update load properties progam to be more flexible that can use either default properties file or any other properties file.
• Add function to keep track what environment the job is running on.

• Develop program to get new fiscal year fund names into Excel sheet.
• Develop program to assign new fiscal year fund allocations:
  • Get fund name, allocation amount, fund id and ledger id from Excel sheet.
  • Use Voyager APIs to get each fund's full data and replace fund original allocation amount to the new amount if this fund original amount is 0.

• Create example programs to test three tomcat realms.
• File-Based Realm: UserDatabase:
  • Add user names and role names in tomcat-users.xml.
  • Test authentication mechanisms in two choices: BASIC and FORM.
  • Add <security-constraint>, <security-role>, and <login-config> in web.xml.
  • Use default realm mechanism in server.xml: <Realm className="org.apache.catalina.realm.UserDatabaseRealm"/>.
  • Combine Yale's CAS netid and password into tomcat realm as login authentication verification, so add dummy password in tomcat-users.xml.
• JDBC Realm:
  • Add user names and role names in operator1(users), roles, user_roles tables.
  • Test FORM authentication mechanisms.
  • Add <security-constraint>, <security-role>, and <login-config> in web.xml.
  • Add JDBC realm mechanism in server.xml: <Realm className="org.apache.catalina.realm.JDBCRealm" connectionName="xxxxxx" connectionPassword="xxxxxx" connectionURL="jdbc:oracle:thin:@<HOST>:<PORT>:<SID>" debug="99" driverName="oracle.jdbc.driver.OracleDriver" roleNameCol="role_name" userCredCol="password" userNameCol="operator_id" userRoleTable="user_role" userTable="operator1"/>.
  • Combine Yale's CAS netid and password into tomcat realm as login authentication verification; so add dummy password in operator1(users) table.
  • Try to use Voyager operator table as tomcat realm users table, but it didn't work, because voyager tables and tomcat realm tables are in different schema.
• DataSource Realm: setting are the same as JDBC realm, except for realm configuration in server.xml:
  • <Realm className="org.apache.catalina.realm.DataSourceRealm" debug="99" dataSourceName="jdbc/authority" userTable="operator1" userNameCol="operator_id" userCredCol="password" userRoleTable="user_role" roleNameCol="role_name"/>.
  • Add "jdbc/authority" into <GlobalNamingResources> through tomcat GUI in admin - Resources - Data Sources.

• Graphics creation is part of high quality web-based applications design. Photoshop can greatly help us to achieve the that goal. In this class, I learned selecting parts of images; moving, duplicating, and resizing images; how to use layers and apply layer effects and filters to create special effects, including lighting and texture effects; how to use painting tools and blending modes to create shading effects; performing adjustments to contrast and color balance.

• Figure out the right Voyger API.
• Learn the event process and how to operate from Voyager client.
• Generate Excel sheet for original allocation to be filled in.

• Finish up Circ Reports, Fund Management Tool, Fund Administration Tool and promote them to production run.
• Work on authentication recognized by non-student, individual netid.
• Update YUL_TOOLKIT to include Circ Reports, Fund Management Tool, Fund Administration Tool and Currency Exchange Rates Estimator.
• Solve the problem that applications having VoyAuth filter can't work from staff front door.

• Analyze filter servlet program and add close function into destroy method.
• Work with ITS to change server start and shutdown process to be ensure that each operation is fully completed before other operation starts.

• Tomcat is an open source that is written in Java to run web application. In this class, I learned tomcat configuration, manager, administrator; deploying and administering a web appplication; tomcat authentication, security, JDBC & JNDI approach; ant script.

• Develop add, update, delete, sort function to access selector, subject, fund mapping table.