| 小洛 的个人资料Paul's照片日志列表 | 帮助 |
|
|
2005/6/3 数据查询--SELECT语句(刚才查找一个问题的时候看到了这个网页,收藏一下,嘿嘿,MSN Spaces真的是一个不错的东东,可以让你任意收藏,呵呵。)
在数据库中,数据查询是通过SELECT 语句来完成的。SELECT 语句可以从数据库中按用户要求检索数据,并将查询结果以表格的形式返回。我们在“Transact-SQL 语言”章节及前面的章节中已经初步接触到了SELECT 语句的一些用法,在本章中将分类讲述其具体用法。 本节讲述SELECT 语句完整的语法结构,这是一个非常冗长、枯燥的过程。读者可以跳过本节,从第二节开始阅读,而将本节作为理解、编写查询语句的语法参考资料。 SELECT 语句完整的语法结构如下: 10.1.1 SELECT 子句
table_name | view_name | table_alias.* 10.1.2 INTO 子句 10.1.3 FROM 子句 table_source 注意:如果在WHERE子句中指定一个值为FALSE的条件,则可以用SELECT...INTO语句来创建一个表名不同,但结构和数据类型均和原表相同的表。 10.1.5 GROUP BY 子句
使用Distinct选项的统计函数,如AVG(DISTINCT column_name)、COUNT(DISTINCT column_name)、和SUM(DISTINCT column_name)等,不能在使用CUBE或ROLLUP选项时使用。 10.1.6 HAVING 子句 10.1.7 UNION 操作符 列的数目和顺序在所有查询中必须是一致的; 10.1.8 ORDER BY 子句 order_by_expression 10.1.9 COMPUTE 子句 10.1.10 FOR BROWSE 子句 XML 2005/4/7 在Windows 2000下安装ViewCVS先决条件 1、安装 如果需要图形化显示cvs树,可以从http://www.akhphd.au.dk/~bertho/cvsgraph/下载CvsGraph 2、配置 启用enscript 启用cvsgraph 3、运行 ii.结合apache 或 借助mod_python模块 Alias /viewcvs/ "C:/viewcvs/www/mod_python/" <Directory "C:/viewcvs/www/mod_python"> [Useful links] Python Win32 Extensions:http://starship.python.net/crew/mhammond/ 如果需要图形化显示cvs树,可以从http://www.akhphd.au.dk/~bertho/cvsgraph/下载CvsGraph 2005/4/1 4/1号会议记录,可不是愚人的哦,我们才不过那么无聊的节日,HOHO....时间:2:30 pm
70fileserver上开了一个设计的folder,设计的内容和模板都在里面 2005/3/23 CrystalClear问题跟踪报表需要继续确认的问题: 需要注意的问题:
---------addition---------- cluster环境台湾搭建也有问题,正在和德国厂商沟通 上周提交过去的报表台湾的工程师确认无法做到。peggy确认只做7天和14天两种周期,而不是天数任意。如果将来需要继续确认之前的方法,就要提交到德国公司,但是经过内部讨论,认为违反了crystalclear的基本设计原则,所以不抱太大希望。 isFailedReport方法确认过,无法在cluster环境中使用,也就是说无法判断cluster中其他node的报表的failed/ok的状态。目前说在6.1的版本中增加判断cluster的功能,正在要求其在6.0的正式版本中就提供。 仍待解决:extend自ReportViewBean的sample code还没有发过来,所以还未研究。 报表上的东西可以松一口气了在线报表和排程报表大致上都有了明确的方向,目前只是需要和GIEA的steven确认两个问题就好了,压力没有那么大了。 记录一下问题: 1,Online:怎样隐藏参数?似乎应该用ReportViewBean的API... 2,batch:method:Cache.isFailedReport(ReportCacheKey) is not implemented for the ClusterCache.什么意思?那么在做cluster的时候会有什么问题和影响? 期望的功能: 可以订制Applet中export的项。 发现的bug: 在MicroSoft VM中print的部分会有错误。 翻页的时候有的时候会丢失一部分数据,不过刷新后就正常显示 2005/1/28 How do we configure cluster report on two computers?How do we configure cluster report on two computers? after that start the i-net Crystal-Clear configuration tool (e.g. double click on CrystalClear.jar) on each of these computers (cluster nodes) and specify the same i-net Crystal-Clear Plus license key in the field "License Key" enable the cluster in the dialog "Cache / Cluster Properties" of the configuration tool Now you can send report requests to both i-net Crystal-Clear cluster nodes and each cluster node will ask the other cluster nodes in the i-net Crystal-Clear cluster if the requested report is in the cache of the cluster node. If not, then one cluster node will execute the report. We recommend to use either harddisk cache or database cache (see dialog "Cache / Cluster Properties" of the configuration tool) instead of the memory cache on all cluster nodes. Crystal-Clear的负载平衡What is the i-net Crystal-Clear cluster? In the i-net Crystal-Clear cluster a cluster node is a computer on that i-net Crystal-Clear is running with the same i-net Crystal-Clear Plus license key as on the other cluster nodes (computers) and on that the property "Cluster" is enabled (see configuration tool). If you have installed i-net Crystal-Clear on two computers, you have entered the same i-net Crystal-Clear Plus license key and you have enabled the property "Cluster" on both then you have a cluster with two cluster nodes (two i-net Crystal-Clear engines that can render the reports). 给报表增加数据库的用户/模式Can I specify the owner/schema of the table at run-time? 让Crystal-Clear使用App的connection poolCan I use a database connection from the application server's connection 关于Crystal-Clear关闭缓存机制的相关信息How can i disable the report cache to get always the current data in the report? 在Crystal-Clear中使用ResultSet的方法,唉,Crystal-Clear的资料也太少了How do I provide data for my report when I don't want to fetch data from a database. Via engine.setData(...), where the data can either be an object array or a result-set. There are two possibilities, you can either set the data from within the checkProperties(Engine engine, ...) call-back or you can write your own database class to provide the data. The essential steps to write your own database class are: create (invent) a dll name; for example foo.dll create a class from a file Foo.java (see the sample (DataWithoutJDBC.java in the jdbc folder of the documentation zip file) and compile that with javac -classpath . DataWithoutJdbc.java Visit crystalclear.properties and connect foo.dll to your Foo.class by adding the following lines (foo.dll maps to symbol FOO which defines the class property Foo (without .class extension): foo.dll=FOO FOO.class=Foo run your report with http://...my.rpt&dll=foo.dll. It will use the data you provide. |
|
|