I found a good page in the Eclipse Help pages for Force.com which explains the architecture of a controller extension and custom controller. These are the steps (in general) that happens when a VIsualforce page is requested, loaded, managed and closed.

1. User requests a page

ONLOAD:
2. The constructors are called for a possible associated controller or controller extension.
3. The assignTo attributes on any custom components on the page are executed
4. Expressions are evaluated.
5. The action attribute on the page tag component is called.
6. Other method calls are made (getter and setters).
7. The resulting HTML is sent to the browser.

POSSIBLE ACTIONS:
8. If the page contains a form component, the view state i updated whenever the page is updated
9. As the user interacts with the page, action methods and getter and setters methods are called as the page contacts the controller as required.

CLOSING:
- When the user is redirected to another page, the view state and controller objects are deleted.

We got a request to modify the page that opens when you want to add a partner for an Opportunity.



Sounds like an easy task to start with. However it turns out to be a bit difficult.
The page is a standard page built into the application and can not be changed. The obvious option would then be to create your own Visualforce page instead and override the New button on the Opportunity related list for Partner to point to this Visualforce page instead. However that is not possible either. You can not override the button nor delete it or or create any new buttons.

We came up with two options. None of them has been implemented at this stage.

#1 Create a custom object
- Create a custom object to use instead of the OpportunityPartner object.
- Remove the related lists for Partner from the Account- and the Opportunity page layouts.
- Create a VisualForce page to represent the new Partner Edit page.
- Create new related lists for the custom object and add to the page layout for this object.
- Add New buttons to the related lists and tie it to the new VisualForce-page.
- Run a data loader job to copy data from the existing OpportunityPartner objects to the new custom object.
- Think twice before deleting any OpportunityPartner or Partner objects. We tried that once and deleted lots of Opportunities.
- Create new reports and remove old reports
- Other dependencies in Apex classes or integrations?

#2 Keep the OpportunityPartner object
- Remove the related lists on the page layouts for the Account and the Opportunity
- Create a VisualForce page to represent the new Partner Edit page.
- Create a button on the Opportunity detail which is tied to this VisualForce-page.
- Create another VisualForce page that should simulate the Opportunity Partner/Partner related list.
- Add this VisualForce-page as inline VisualForce to the Opportunity detail.
- Since the OpportunityPartner object is only queryable and retrievable there might be a problem with the Visualforce Page simulating the Partner Edit page. In this case the Partner object could be used instead as the standard controller since a creation of either of the objects creates an instance of the other.

Option #1 is more clean and future safe. However it is more expensive in terms of time and also contains more risk of missing something.

Option #2 works good enough and is not as expensive. Doesn't feel too good to move away from standard behaviour on a standard object though and it is not as future safe.

Since Google Buzz started there is suddenly a need for managing your Google profile. Your profile is more or less visible and clickable depending on how well you have entered your personal information. You have to open your profile and add information and also (optional) choose a username for your profile URL (otherwise it is just a long number). Get your profile by logging into Google at www.google.com and click on 'Edit profile'. You can also goto http://www.google.com/profiles/ and search for your name.

This is how a profile can look like (my own).


You can via Buzz in Gmail search for other Buzz users to follow. Just click 'add' if they seem interesting. You are probably already following many of your Gmail friends. When looking at the people you follow you can click their profiles (if these are enabled) to see more information. It is also possible to go directly to other peoples profiles and follow their buzz from there. A profile's url is on the syntax http://www.google.com/profiles/username.

I have seen several offical profiles without any information which doesn't look too professional.

Check this site for more information, tips and trix about Buzz.

Here are a couple of interesting ones that I have found so far:
- Mashable
- BuzzUsers
- Google Buzz Team
- TechCrunch

I was modifying a test class for a trigger when I bumped into this error message - "List has no rows for assignement to SObject". This means that an object instance that is to be assigned a value from a SOQL query does not get assigned, i.e. the SOQL query return no value.

My case was something like this:

Opportunity opp = [select id, StageName where id = :someId]

If the SOQL query for some reason would not a return a result row the error message would be raised. A way to solve this is to assign the result of the SOQL-query to an Opportunity array instead.

Opportunity[] opps = [select id, StageName where id = :someId]

and then check with conditions:

if (opps.size() > 0) { /* At least one row has been returned from the SOQL query If we for example are only expecting one opportunity as result then get a handle on it like this */ opps[0].CloseDate = date.today(); } else { // Query was empty } update opp[0];

This way no error is raised.

Now you can create and save a PDF for any website for free with PDFmyURL.com either by typing in the url by hand on the site or by calling the website url with querystring parameters via a link or directly in the address field of the web browser.

This link:
Download a random wohill page as a PDF

would look like this in HTML:

<a href="http://pdfmyurl.com?url=www.wohill.com/random">Download a random wohill page as a PDF</a>

As you can see the whole blog framework with one page is created as a PDF. It could be interesting here to point to just the blog entry and being able to offer random articles as PDF-pages for example. That will be a smaller project for a rainy day...

There are also advanced options for the PDF such as specifying the filename, choosing landscape or portrait, adding text, adding header or footer etc. Read more on the first page of PDFmyURL.com.

Interesting tool. I am not sure if I like it or not. It doesn't look too good for most sites though.
I was aware that some people still uses a low resolution but I didn't think it was this bad.

However I interpret it the way that they have taken into account also parameters like when visitors don't have their browser window expanded to the maximum or when visiting with a mobile device (and the site is not specifically enabled fior that).



Something to consider anyway. Check it out for your own site!

If you have created an Apex class which uses Attachments in some way you might need to create Attachments as test data for the testMethod in order to get the coverage to be able to deploy.

The attachment has three required fields: Body, Name and ParentId
But how is the body, which is the attachment, created programmatically?

I found this solution in a Salesforce community and also later on the Salesforce blog:

PageReference pdf = Page.testPDF; pdf.getParameters().put('p','p'); pdf.setRedirect(true); // Grab the PDF! Blob b = pdf.getContent(); // Create and insert attachment Attachment attachment = new Attachment(Name='An attachment',body=b,parentId=accId); insert attachment;

The above assumes that you have a valid accountId 'accId' serving as the parent of the attachment and that you have created a Visualforce page 'testPDF' which is rendered as PDF. Could look like this:

<apex:page renderas="pdf"> <!-- Begin Default Content REMOVE THIS --> <h1>Congratulations</h1> This is your new Page <!-- End Default Content REMOVE THIS --> </apex:page>

Notepad++ is probably the best developer tool you got when it comes to web programming - Clean and simple.

I recently installed a new plugin for the BigMachines BML language - read more about how to do that here. Works great by the way!

Today when I was creating a new library function for the commerce process I started to annoy myself that the color of the single line commenting were so much brighter than the block comment and hard to read. Now, instead of keep being annoyed we could do something about it and it's very simple.

- Open the file ' userDefineLang.xml' which should be located in %APPDATA%\Notepad++.
- There's a tag named 'Styles' which child tags are named 'WordsStyle'. The one with name 'COMMENT LINE' is the one I want to change the color of. So I just copy the color of the block comment tag (with name 'COMMENT') and use for the line comment tag instead. Should look like this:

<WordsStyle name="COMMENT LINE" styleID="2" fgColor="808080" bgColor="FFFFFF" fontName="" fontStyle="0" />

Save and restart and we're done!

Did you know that you can read Twitter statuses tagged with hashtags directly on Google? That is if you're using the Chrome web browser.

Search for the words 'Twitter' and any valid hashtag and you will be presented with a scrollable html div which displays the latest tweets with this hashtag. It doesn't seem to update in real-time though.

It is tempting to do it like this:

document.getElementById('idOfTheDiv').style.width = newWidth;

However this won't work and if you think about it, it is not strange either.

This is how to declare the width and height of a div in a stylesheet:

aDiv { width: 200px; height:100px; }

And by looking at this we can see why the first example doesn't work and how to fix it. I.e. add 'px' when setting the width like this:

document.getElementById('idOfTheDiv').style.width = newWidth + 'px';

Here is a simple example of how to easily achieve Ajax-functionality with Domino and Ext.js.
The idea is to, from a page that is opened in a web browser, call a Domino agent asynchronously and in the background to do some work and to provide feedback on it. The page is then updated when the work is done.

1. Download the Ext core (http://www.extjs.com/products/extcore/download.php). This is a JavaScript library that among others contains functionality that can be used to achieve Ajax functionality easily.

2. Copy the contents from the JavaScript file 'ext-core.js'

3. In your Lotus Notes database, create a Javascript library named ext-core.js and paste in the copied content from step 2.

4. Create a page with some Pass-Thru HTML.
- A button which when clicked on runs a JavaScript function 'runAjax();'
- A div-tag in where the result from the Ajax query is displayed

<table> <tr><td><input type="button" value="Run" onclick="runAjax();" /></td></tr> <tr><td><div id="results"></div></td></tr> </table>

5. Add a script tag in the 'HTML Head Content' section of the page to include the JavaScript library to the page.

<script type='text/javascript' src='ext-core.js'></script>

6. Define the Javascript function 'runAjax()' in the 'JS Header' section of the page. This function calls an url, in this case a Domino agent, which is run. If the agent was run successfully and the call and response was successful we can decide what to do. Here we update a div on the page with the response from the agent, i.e. what the agent prints.

function runAjax() { Ext.Ajax.request({ url: 'myController?openagent', method: 'GET', success: function(response, opts) { document.getElementById('results').innerHTML = response.responseText; }, failure: function(response, opts) { alert('server-side failure with status code ' + response.status); } }); }

7. Create the agent 'myController'.
The idea with the agent is to have it run the actual logic and that would be work with Domino data for searching or modifying for example. The agent could just do something with data but also send back information or status to the page that called it. In this example the calling page only runs the agent and the agent only prints a text line as a response or feedback.

Sub Initialize() Print "Foo bar" End Sub

8. Done! Open the page in the web browser and click the button. The text 'Foo bar' should be displayed shortly in the div.

Note: This does not work in the Notes client, only in a web browser.

SOQL (Salesforce Object Query Language) is the language used in salesforce when writing Apex classes for example to retrieve sets of data.

SQL (Structured Query Language) is the language for managing data in relational database systems.

The two query languages are very similar but differ in different distinctive ways. In general SOQL is not as powerful as SQL and don't have all the same functions. Here are three of the more imporant differences as bullet points:

- SOQL is only used to retrieve sets of data while SQL can directly retrieve and modify sets of data.

- SOQL can retrieve data from related Salesforce objects using extended dot notation. However the linking only works for pre-existing relationships. In other words SOQL can't be used to join unrelated sets of data while that is possible with SQL.

- SOQL does not support the full set of SQL syntax. Some examples:
- 'SELECT * FROM Account'
- 'SELECT DISTINCT id FROM Account'.
-SELECT Id, Name FROM Account GROUP BY Name



Please feel welcome to add other differences or more examples of the limitation of syntax in relation to SQL.

If you develop in Eclipse you change it in the meta data part of the class.



If you develop directly in the web browser, you change it in the version settings tab.



Each new version of Salesforce consists of a new version of the API which includes richer features and greater efficiency as it says in the help. So to be able to use new functionality you much specify a more recent API version. However, the API versions are backwards compatible.

I wanted to create a simple XPages application for a specific purpose but since I haven't learned that yet and didn't have the extra time to do so either I had to go for the old style version. So I created a web form with some JavaScript and CSS. What I explain below is probably old news for most of you but I wasn't aware of this so I'm sure there are others.

I created a form with passthru HTML and added combo boxes to be able to select values from. The values were taken from a DbLookup from a keywords view. This looked fine and also worked fine in Firefox and Google Chrome. But for Internet Explorer (7) the values wasn't passed when clicking save and calling the submit function.

So I checked the source code and saw that the options lacked values. For Internet explorer the select box was built like this:

<select> <option>Value 1 <option>Value 2 <option>Value 3 </select>

And hence no values were passed.

For Firefox and Chrome however the select box was by default built this way:

<select> <option value="value1">Value 1</option> <option value="value2">Value 2</option> <option value="value3">Value 3</option> </select>

The workaround:
Since I didn't need to save the values chosen but only needed them to save to other fields I could instead print HTML in a computed text that allowed the select box to print correctly also for IE.

Here's the formula code for that:

tmp := @DbLookup(Notes:Recache;"":"";"LookupKeywords";"Keyword";"Values"); str := "<select id='someId' class='comboboxCssClass'>"; str := str + @Implode("<option value='" + tmp + "'>" + tmp + "</option>"); str := str + "</select>"; str

I am currently learning Salesforce. I have attended the two developer trainings available, DEV-401 and DEV-501 and in my current daily work there are both easier and more complex assignments for me to take care of which gives me a good chance to learn more.

So please bare with me if you find some tips and entries a bit basic in the beginning. But since I had questions I am sure someone else do to.

Now to the point of today's entry.
I recently blogged about how to prevent a delete button from delete under special circumstances. I tested this and some others did too. What I didn't think of was that all of us that tested were system administrators so we didn't see what I had missed.

If you have created a visualforce page and use that to for example override a button, all users will not be able to access this page. This is because you will have to add this to the profile(s). This is true for both Apex classes and for Visualforce pages.

What you do to make this work is:
- Goto setup -> Manage Users -> Profiles and do the following for all profiles that need to see the visualforce page or apex class.
- Scroll down to the sections 'Enabled Apex Class Access' or 'Enabled Visualforce Page Access'.
- Edit either of them and add from the available list to the enabled list and save.

The System Administrator Profile automatically has access.

I recently had a case where I needed to remove the possibility to delete a Product (child to the Opportunity).

1. The first obvious solution would be to remove the delete access for the users. However since we are using a third-party product configurator to integrate with Salesforce this was not possible since they used the user rights to remove products from their system.

2. Second I though of creating a Visualforce page to mimic the Opportunity, i.e. using the Opportunity as standard controller and then overriding the related list which is the place where you can delete using the 'Del' link. However, there seems to be limitations on what you can do here when working with custom related lists.

3. Third to keep the 'Del' link but prevent the system from deleting by overriding the delete links och delete buttons for the Product object. Currently this is my best option.

3.1 Create a Visualforce page which says something like "You are not allowed to delete Opportunity products" and provide a link to the previous page.

3.2 Override the delete link on the Opportunity Product object with this VF-page.

At first It still seems as if the user can delete but when he clicks on the delete link the confirmation window first pops up and after that this new Visualforce page opens up and no deletetion has occurred (which is the most important).

Here is the Visualforce code. I have used static resources as you can see:

<apex:page standardController="OpportunityLineItem"> <apex:stylesheet value="{!$Resource.customStyle1}"/> <apex:includeScript value="{!$Resource.curvyCorners}"/> <apex:includeScript value="{!$Resource.curvyCornersSrc}"/> <apex:includeScript value="{!$Resource.customJavascript1}"/> <div style="width:100%;" align="center"> <br /><br /> <div id="userRoundedCornersBox"> <h1>FORBIDDEN!</h1> <br/><br/> You are not allowed to delete Opportunity products from Salesforce. This can only be done from the product configurator. <br /><br/> <a class="vfdp" onclick="goBack();">Click here to go back to the previous page!</a> </div> </div> </apex:page>

And this is how it looks like:


Do you have a better suggestion? Please leave a note in the comments.

Todays blog entry will be an open question to any XSL-developer.

Is there a way to use the HTML equivalent 'rowspan' in XSL-FO? I want to display a text with a bigger font in the leftmost column. The colums to the right however should consist of a couple of rows with smaller text.

In HTML this would be written something like this:

<table border="1" cellpadding="4" cellspacing="0"> <tr> <td rowspan="3" style="font-weight:bold;font-size:14pt">New Text</td> <td>Value 1</td> <td>Value 2</td> <td></td> </tr> <tr> <td>{someAttribute}</td> <td>{someOtherAttribute}</td> <td></td> </tr> <tr> <td>Value 3</td> <td>Value 4</td> <td>Value 5</td> </tr> </table>

and look like this:

Here is the code for XSL-FO without the new leftmost column.

<xsl:template name="printAddress"> <xsl:param name="subDocument"/> <fo:table table-layout="fixed" background-color="#e0e0e0" keep-with-next.within-page="always"> <fo:table-column column-width="7.0cm"/> <fo:table-column column-width="7.0cm"/> <fo:table-column column-width="2.0cm"/> <fo:table-body> <!-- Begin Row 1 --> <fo:table-row keep-with-previous="always"> <fo:table-cell margin-left="0.2cm" padding-before="0.1cm" padding-after="0.1cm"> <fo:block>Value 1</fo:block> </fo:table-cell> <fo:table-cell margin-left="0.2cm" padding-before="0.1cm" padding-after="0.1cm"> <fo:block>Value 2</fo:block> </fo:table-cell> <fo:table-cell margin-left="0.2cm" padding-before="0.1cm" padding-after="0.1cm"> <fo:block/> </fo:table-cell> </fo:table-row> <!-- Begin Row 2 --> <fo:table-row keep-with-previous="always"> <fo:table-cell margin-left="0.2cm" padding-before="0.1cm" padding-after="0.1cm"> <fo:block> <xsl:value-of select="$subDocument/someAttribute"/> </fo:block> </fo:table-cell> <fo:table-cell margin-left="0.2cm" padding-before="0.1cm" padding-after="0.1cm"> <fo:block> <xsl:value-of select="$subDocument/someOtherAttribute"/> </fo:block> </fo:table-cell> <fo:table-cell margin-left="0.2cm" padding-before="0.1cm" padding-after="0.1cm"> <fo:block/> </fo:table-cell> </fo:table-row> <!-- Begin Row 3 --> <fo:table-row keep-with-previous="always"> <fo:table-cell margin-left="0.2cm" padding-before="0.1cm" padding-after="0.1cm"> <fo:block>value 3</fo:block> </fo:table-cell> <fo:table-cell margin-left="0.2cm" padding-before="0.1cm" padding-after="0.1cm"> <fo:block>Value 4</fo:block> </fo:table-cell> <fo:table-cell margin-left="0.2cm" padding-before="0.1cm" padding-after="0.1cm"> <fo:block>Value 5</fo:block> </fo:table-cell> </fo:table-row> </fo:table-body> </fo:table> </xsl:template>

The HTML-equivalent for the above code would be to only remove this row from the provided HTML code at the top:

<td rowspan="3" style="font-weight:bold;font-size:14pt">New Text</td>

How would the XSL-FO code look like to provide the same look as the figure at the top?

Boxes or divs with rounded corners can be nice to present content in, but how to do it?
Well, there are several ways using both CSS and JavaScript and here are just a few of them.

My impression when looking into this is that the javascript frameworks/toolkits don't have these functions built-in easy to use but instead you will have to use one of the below methods yourself.

- RoundedCornr (it is done for you)
- CurvyCorners (A free JavaScript library)
- SpiffyCorners (done for you)
- Rounded Corner Boxes the CSS3 Way
- CSS rounded corners
- Advanced CSS rounder corners
- Rounded corners reference (CSS and JavaScript)

Please feel free to complement with any info or links. I was looking into this for a specific purpose and wanted to know of the best solution to go for.

I have been taking it slow with Notes and Domino for the past year because of other obligations. Now I got the opportunity to create a webapp and I decided to do it in Lotus Notes and I thought I should go for Notes 8 and XPages to get a nice feeling for the user. I have to develop it in a pretty short time frame as well.

So I need to search and browse through some material for documentation and examples to get started. Do you recommend any special source on documentation, references or examples regarding XPages that I could use?

Any lessons learned or other info that might be good to know about since its been out for some time now?

I have specified some links below that seemed useful to me when Googling it quickly.
Appreciate any links on this!

- IBM Knowledge Collection: XPages
- The XPages Blog
- Learning XPages By Declan Lynch
- XPages in Lotus Domino Designer Wiki
- Harness the power of XPages in Lotus Domino Designer

When I download a file from one of the data tables in BigMachines I want to modify it in some way and then upload it back again.

If there are hundreds or thousands of rows I need this data manipulation to be in Excel. Sometimes there are Swedish characters like å, ä or ö in the file and they need be be preserved this way.

However when opening a csv-file encoded with UTF-8 in Excel these characters are displayed as very odd characters instead. If I replace the odd characters with the correct ones and save it, it is saved as ANSI and uploading it will instead make it look strange and cause errors in the system.

Here is a little workaround to make it work.
1. Open the csv-file in Notepad++ and convert it to ANSI and save.
2. Import to excel and modify and save.
3. Open again in Notepad++ and convert to UTF-8

A Salesforce id represents a specific object instance in Salesforce. For some objects the id starts with a specific set of characters (prefix) in the beginning of the id string, see here.

You can obtain the Salesforce id in different ways, either directly in the Salesforce application or by using the API.

1. The most common and easy way is to look in the address field of the browser. This provides you with the 15-digit Salesforce id which is also case-sensitive.



2. You can also get the Salesforce id by using the Salesforce API like Apex Explorer or Data Loader. This provides you with the 18-digit Salesforce id which is not case-sensitive. This way it can be processed by Excel without a problem for example.



Use whichever you want, they both work in all scenarios what I know of.

If you for some reason need to convert a 15-digit id to a 18-digit id there's an algorithm for that.

Salesforce provides different tools for reading and/or writing data from and into Salesforce. Two of these are the Data Loader and Apex Explorer.

For both of these you have to specify username and password when starting the application in order to use it. The username is the Salesforce username but the password for these two tools are the Salesforce password + Security token. This is what's used for API-based communication in general with Salesforce.

However there is a way to get away from having to think about the security token and that is to have Salesforce put your IP address as a trusted one. Once you have done that you will only have to specify the Salesforce password in the dialog box.

You do this by going into the Network Access settings. Locate it in the setup under 'Administration Setup..


Then add your IP-address as shown below.

Salesforce provides podcasts on developer training available on iTunes. Just search for Salesforce in the iTunes Store and you will find the podcast salesforce.com Training and Certification. On this podcast the entire dev-401 and dev-501 courses are available as videos. There are also a Admin to Hero with Force.com module including explanatory videos of the Salesforce.com admin GUI for creating for example a validation rule.

If you don't have iTunes or if you simply are not allowed to use it if you are behind some corporate firewall, you can download the files directly from http://www.salesforcetrainingpodcasts.com. See the dev-401 directory, the dev-501 directory and the Admin to Hero directory.

Anyway, it is really great that this has been made available by Salesforce.

Salesforce has many standard objects. I am not sure if this is true for all, but the most common ones at least have fixed prefixes which could be good to know at least for programming purposes I can imagine. Here is a list of the most common objects and their prefixes.

Prefix		Standard Object
001 002 003 005 006 00Q 500 00T 00U 015 800 00P		Account Note Contact User Opportunity Lead Case Task Event Document Contract Attachment

Is it possible to copy Salesforce object instances and preserving integrations with other systems?

Let's say you have a production environment in Salesforce with lots of data. Now you are either creating a sandbox or refreshing an existing sandbox with the production environment.

If you want to bring your data or part of your data for testing purposes you can do this by manually exporting the objects from the production environment object by object and then uploading them to the sandbox - using the Data Loader. This is time-consuming if you have many objects with different relationships to each other - but it will end up right in the end.

! Please let me know if there are easier ways of if someone has written some kind of script for doing this using the Salesforce WebServices API for example.

Now, if you have an integration from Salesforce to a configurator, for example BigMachines, the configurator will contain configurations as well. With BigMachines, these configurations refers back directly to Salesforce Id's. So for refreshing a Salesforce environment and preserving data means that you have to take into account that the integration works afterwards as well. To be honest this is just not a configurator or BigMachines issue but really any integration.

Taking this further means that in order for the sandbox to work as the production environment does, the BigMachines configurations and the Salesforce objects still have to be connected. Since BigMachines refers to specific Salesforce Id's we need to move the Salesforce instances from production to the sandbox preserving the same id's.

This is not possible!

I have tested by extracting an Account from our production environment. Removing all references to other objects not needed and just keeping pure data and the Salesforce Account Id.

An insert of this row into the sandbox ignores the Salesforce Id and creates a new.
An upsert into the sandbox demands me to choose a matching id which is either the Salesforce Id or an external id. When choosing the Salesforce Id I get an error saying "invalid cross reference id". If I choose an external id as the matching id I get an error saying "cannot specify Id in an insert call".

Now, this seems logical to me that it doesn't work.
Imagine you have a MySQL or any relational database with a primary unique id in one of its tables. An insert here would behave the same way and an update on an id that doesn't exists in the table would not be possible. An update on another id would be possible but fail it would try to set the id at the same time.

But the reason that I bring this up is the need to move data with consistency with integrations taken into account. I guess the only solution would be to make the integrations look up on external id's instead - id's that are not primary and auto-incremental in the same sense and that are preserved during a copy. That would raise other risks and issues I guess but those could probably be handled

Worth mentioning though, is that
It's possible to copy the exact environment with the exact data where integrations still work and id's are preserved if you make a complete refresh.
In our case the sandboxes don't hold that storage capacity so that's why it's not possible.

This might be useful to somebody.

Let's say you have a custom object which refers to and is a child object to the Opportunity object called OpportunitySubObject.

In Apex Explorer you want some values from both the Opportunity and from OpportunitySubObject based on some criteria. As you might know you just click in the fields that you want to list and you add the condition yourself. The given select statement becomes something like below. As you can see the select statement contains both values from the Opportunity and from the OpportunitySubObject:

select o.Opportunity.Number_of_Something__c, o.OpportunityId, o.Id from OpportunitySubObject o where o.Opportunity.Number_of_Something__c = 0 and o.ownConditionField__c = 'condition x'

Now, if you want to export the result of this query using the Data Loader it won't work. You will have to modify it slightly by removing the references to OpportunitySubObject in the select part:

select Opportunity.Number_of_Something__c, OpportunityId, Id from OpportunitySubObject o where (o.Opportunity.Number_of_Something__c = 0 and o.ownConditionField__c = 'condition x')

I recently blogged about how to display a text as columns with PHP.
Now, thanks to PHP.JS, which I have also blogged about, I have ported the same functionality to JavaScript.

No work from my side really other than using the corresponding PHP functions from PHP.JS and creating my own JavaScript split_pos function. I have also created a print_columns function that calls split_pos and renders the columns on the web page. Currently, as you can see, this function is pretty hard-coded and can only handle one and two columns. I will update this entry with a more abstract version if I get the time to do so.

You need to download and include the package from PHP.JS for this to work since several ported JavaScript functions from this library are being used.

function split_pos(text) { var mid; var cut; var part1; var pos1; var pos2; var pos3; /* find middle space in text */ mid = strlen(text)/2 - 1; cut = strpos(text , ' ' , mid); part1= substr(text , 0 , cut + 1); pos1 = strrpos(part1 , '<'); pos2 = strrpos(part1 , '>'); if ((pos1 < pos2) || (pos1 === false)) { return cut; /* no html tag around */ } pos3 = strpos(text , '>' , cut1 + 1); if (pos3 !== false) { return pos3; /* end of middle html tag */ } else { return cut; /* unbalancing < > */ } } function print_colums(text, column_spacing, column_width, columns) { var mid_pos = split_pos(text); var col_str = ""; col_str = "<table cellpadding='0' cellspacing='0'><tr valign='top'>"; if (columns == 1) { col_str = col_str + "<td class='textContent' style='width:" + column_width + "px'>" + text + "</td><td style='" + column_spacing + "px'></td><td class='textContent' style='width:" + column_width + "px'></td>"; } else if (columns == 2) { col_str = col_str + "<td class='textContent' style='width:" + column_width + "px'>" + text.substr(0,mid_pos) + "</td><td style='width:" + column_spacing + "px'></td><td class='textContent' style='width:" + column_width + "px'>" + text.substr(mid_pos + 1) + "</td>"; } col_str = col_str + "</tr></table>"; return col_str; }

Sometimes it happens to me that I really would like to use a PHP function in JavaScript. Instead of making unnecessary AJAX calls to accomplish this it would be nice if you could use the same functionality in JavaScript.

Now, this happens to be possible in an easy way thanks to PHP.JS, which is an open source project in which PHP functions are ported to JavaScript functions.