Translator – Reference Flag Implementation

For non-parentheses tokens, only the NoParen and DefFuncN token types should have the reference flag set. The NoParen token type includes a possible variable or user function with no arguments. The reference flag should not be set for internal functions with no arguments (IntFuncN) or constants. At the moment it's unclear if the reference flag should be set for defined functions with no arguments (DefFuncN). Intuition says that the reference flag should be set since the DEF command line is similar to an assignment statement.

For parentheses tokens, only the Paren (an array or user function with arguments) and DefFuncP token types should have the reference flag set. For now the reference flag will be set for define functions with arguments (same reason as above). The reference flag should not be set for internal functions with arguments (IntFuncP). However, with that said (and this is jumping ahead a bit), some internal functions, namely string functions, could be references. Specifically the MID$ function, which could be used to assign a part of a string. This will be implemented later when data type handling and string handling is added.

Lastly, in the test code that outputs the tokens, a “<ref>” will be output after the token if the reference flag of a token is set. This check will be made for all token types just to make sure it is not getting set when it's not supposed to.

Translator – Single Vs. Multiple Assignments

Some issues came up during the implementation of the assignment handling. A new Assign code is needed for assignment operator. The original plan was that this operator would handle both single and multiple assignments. At run-time, it would simply keeping popping references off of the evaluation stack and assign the value until the stack is empty. However, this requires a check if the stack is empty. The goal is to reduce the number of checks that is needed at run-time because the more checks made, the more time wasted doing these checks during program execution.

Since a single assignment will be the common case and multiple assignment the exception, the single assignment execution time would be penalized by the extra stack empty check To prevent this, two assignment codes are needed, Assign for single assignment and AssignList for multiple assignment.

To implement two assignment codes in the Translator, at the first equal, the token will be changed to Assign as previously planned. If the mode is Equal (indicating another equal in a multiple equal assignment), the Assign token on top of the hold stack will be changed to AssignList. If the mode is Comma, then the token will be changed to AssignList before pushing the token onto the hold stack. At the time the assignment token is added to the output list, the Assign operator will expect two operands (the first a reference) and the AssignList operator will expect multiple reference operands and the value to assign operand.

This brings up another issue. The Recreator will need to distinguish if the AssignList is from a multiple equal or comma separated multiple assignment so that the original code is recreated as it was entered. There could be two different AssignList codes for each (for a total of three assignment operators). Though not implemented yet, there will also need to be unique codes each assignment if the optional LET command word is entered (that's now 6 unique assignment codes). All these unique codes would not be required at run-time (only two are needed, single or multiple). There's actually a better way to prevent this multiplying of codes, but this will wait until the internal code is developed (and it will also eliminate the need for the dummy parentheses codes in the program, but not all the code needed to determine when it is needed).

Translator – More Mode Processing

There is some mode mode processing needed. When a non-assignment operator is added to the output, the mode will determine how the operator token is processed:

Command: An operator is not expected so an “unexpected operator” error occurs.

Equal: Within an assignment (possibly multiple). The mode will be changed to Expression indicating the start of the expression to assign.

Comma: Within a comma separated list of a multiple assignment, an operator is not expected so an “expected comma or equal” error occurs.

Expression: Within an expression, operator is processed as previously implemented.

At the end of the line, some checks needs to be made based on mode to see if any error has occurred:

Comma: Within a comma separated list of a multiple assignment, but there was no assignment operator, so an “incomplete assignment” error occurs.

No errors are detected for the other modes at the current time. Now on to implementation of assignment statements...

Translator – Comma Token Processing

Some comma token processing has already been implemented – commas separating subscripts of arrays or arguments of functions. Commas outside of arrays or functions caused an “unexpected comma” error. This processing needs to be extended to include comma separated multiple assignment statements. Again the current mode will determine how the Translator processes comma operator tokens:

Command: The beginning of a comma separated multiple assignment. The mode will be changed to Comma indicating a comma separated multiple assignment. The comma token is not needed, so it can be deleted. Eventually an assignment operator will be pushed on the hold stack.

Equal: Continuation of a multiple equal assignment statement, so a comma token at this point cause an “unexpected comma in assignment” error.

Comma: Continuation of a comma separated list of a multiple assignment. No further action is needed so the comma token can be deleted.

Expression: Within an expression, so proceeds with the previously implemented comma processing. When the counter stack is empty or the top counter has a zero value, an “unexpected comma in expression” error occurs (changed from “unexpected comma” to differentiate it from the “unexpected comma in in assignment” error).

The comma token has the same low precedence as closing parentheses and assignment operators so most tokens will be removed from the hold stack. Comma tokens are not pushed to the hold stack so they won't be removed. Closing parentheses are also not pushed to the hold stack. And there will never be an assignment operator pushed to the stack before a comma token is processed.

Translator – Equal Token Processing

The Parser returns an equality operator token for an equal. The current mode will determine how the Translator processes equality operator tokens:

Command: The token will be changed to an assignment operator and the mode will be changed to Equal indicating a possible multiple equal assignment and to prevent a comma separated assignment. The assignment operator will be pushed onto the hold stack.

Equal: Continuation of a multiple equal assignment statement. There is already an assignment operator on the hold stack, so this token can be deleted. No further action is needed. During run-time, the single assignment operator will handle multiple variable references.

Comma: The end of a comma separated list of a multiple assignment. The token will be changed to an assignment operator and the mode will be changed to Expression indicating the start of the expression being assigned. The assignment operator will be pushed onto the hold stack.

Expression: Within an expression, equals are equality operators, so the token (already an equality operator) will be processed as a regular binary operator.

The assignment operator will have a very low precedence, which will keep on the hold stack until the end of the statement. Its precedence can be the same as the closing parentheses. The assignment token will be first on the hold stack, so there won't be a closing parentheses token to remove. The closing parentheses only empties the hold stack to an opening parentheses or internal function token (with a lower precedence), so there won't see an assignment token.

Translator – Multiple Assignment Vs. Equality

There are two types of multiple assignment statements being supported – multiple equals and comma separated. Having a two value “in command” flag is sufficient for determining whether an equal is an assignment or equality operator. However, it's not sufficient for knowing which type of multiple assignment is present. For this, two more values are needed, whether in a multiple equal assignment or in a comma separated assignment.

Instead of calling this the “in command” flag, it will be called mode. There will be four modes: in command, in multiple equal assignment, in comma separated assignments, or in expression.
In addition to knowing if the Translator is in command mode or expression mode, the Translator also needs to know if it is in a multiple equal assignment or a comma separated assignment. At the beginning of the line, the mode will be set to command.

For equal tokens, comma tokens and other operator tokens; this mode variable will be used to determine what action to take or what error needs to occur. One of the error conditions to detect is if the two types of multiple assignment statements are mixed, something that is not allowed.

Translator – Assignment Vs. Equality

The Translator needs to decide when to add an equality operator to the output list and when to add an assignment operator. For now, commands won't be considered since support for commands has not been added to the Translator yet. Consider these assignment statements:

A = X * 5
A = B = C + D
A = B + (C = 5) * D
A = B = C + (D = 5)

There needs to a flag for when an equal should be an assignment operator and when it should be the equality operator. At the beginning of a line, a command is expected, so when there is an equal, it should be interpreted as an assignment operator (which is technically an assignment command). Once the expression starts, any equals should be interpreted as equality operators.

In the second example statement, the first and second equals are assignment operators since the expression to be assigned hasn't started yet. In the third example, the plus begins the expression, so the second equal is an equality operator. In the last example, the plus again begins the expression, so the third equal is an equality operator.

The flag will indicate whether whether the Translator is within a command and will be initialized to on at the beginning of the line. If a non-assignment operator is added to the output list this “in command” flag will be turned off. At each equal, if the flag is on then an assignment operator will be put on the hold stack; and if the flag is off, then an equality operator will be put on the hold stack.

Translator – References

When processing operands, the Translator needs to distinguish between the values for and references to variables and arrays. This will be accomplished with a flag in Token to indicate whether it contains a reference or not. When a variable or array token is added to the output list, it's reference flag will be set. When the Encoder processes tokens from the output list to generate the internal code, it will generate a push reference instruction if the reference flag is set and a push value instruction if the reference flag is not set.

In the Translator, when a non-assignment operator is processed and it pops it's operands off of the done stack, it will clear the reference flag in the token of the operand since it needs values and not references. For an assignment operator, the reference flag of only the second operand will be cleared; however, the reference flag of the first operand needs to be checked to make sure that it is set, so that expressions like “5 = A” or “A+B = 4” will cause a “non-reference cannot be assigned” error.

As previously mentioned, the Translator does not know the difference between a variable and function with no arguments or an array and a function with arguments. Therefore, it could be setting the reference flag for a token that refers to a function and not an variable or array. This will not cause a problem however. Only tokens on the left side of an assignment operator will have the reference flag set and there should be no functions on the left side (except in this case of setting the return value of a function with it's own function name – therefore setting the reference flag makes sense, but functions are much later).

Assignments and References

Assignment statements will be processed like any expressions with binary operators, however its first operand is handled differently than the second operand (which is the same as operands of other binary operators).  Consider the expression with its RPN translation:

A + B  A B +

During run-time, the value for the variable A is pushed onto the evaluation stack followed by the value of variable B. When the add operator is executed, it pops the two values off of the stack, adds them, and then pushes the resulting value back onto the stack. Now consider an assignment expression with its RPN translation:

A = B  A B =

During run-time, the assignment operator needs to know where to store the value being assigned. So, for the A operand, a reference to the variable value needs to be pushed onto the evaluation stack instead of its value. The assignment operator will not be pushing anything back onto evaluation stack.

The assignment operator will always be the last operator to be processed, so to handle multiple assignments, the multiple references will be pushed onto the evaluation stack. The assignment operator will simply keep popping references and assigning the value until the stack is empty.

Assignment Statements

In BASIC, the “=” character represents both the equality operator and the assignment operator – it all depends on where the “=” is found. The is different from C, which has unique operators for equality (“==”) and assignment (“=”). Consider these examples:

IF A = 5 THEN
A = 5
A = B = 5

The first example contains equality operator and the second example has the assignment operator. In the third example, does the statement contain two assignment operators (like C) or does it contain an assignment operator and an equality operator? In GW-Basic, SmallBASIC, FreeBASIC and QuickBASIC, it is the latter assigning A to a true or false value depending on where B is equal to 5 or not.

Multiple assignments can be convenient so they will be supported like in the third example above. But it can also be convenient to be able to use the equality operator (or any relational operator) in expressions beyond conditional expressions (in IF, WHILE, UNTIL, etc. statements). Two forms of multiple assignments will be supported:

Variable1 = Variable2 = Variable3 = Expression
Variable1 , Variable2 , Variable3 = Expression

For the first form, the “=” characters will interpreted as assignment operators until the expression starts, in other words, once there is a another operator including an expression in parentheses:

Variable1 = Variable2 = (Variable3 = Expression)
Variable1 = Variable2 = Variable3 + Expression1 = Expression2

Here Variable1 and Variable2 will be assigned to a true or false value depending on whether Variable3 is equal to the Expression in the first example of Variable3 + Expression1 is equal to Expression2 in the second example (the third “=” in both examples is the equality operator).

Translator – Internal Functions (Release)

During testing of the “wrong number of arguments” error, it was noticed that the error was reported on the closing parentheses. While this is acceptable, it would be better if the error pointed to the actual function. However, the caller always pointed to the token that caused the error. Therefore, in order to report an error at any token, the token pointer argument of the add_token() function was changed to a reference, so that it could be changed to any token so that caller will point to this token instead when the error is reported.

The name2 member is being used for the name to output during testing, so for example, for Mid2_Code and Mid3_Code, name2 is set to “MID2$(” and “MID3$(” so that is can easily be seen which code is in the token. This was needed when outputting internal function tokens in addition to operator tokens. So instead of copying these lines of code, a new access function debug_name() was added to get the appropriate name for debug test output.

I also realized there were some memory leaks in the code - the EOL token was not deleted; when an error occurs, the token of the error was not deleted; and when the token is replaced (see above), the original token was not deleted. Therefore, the code was cleaned up to prevent these memory leaks.  Nothing was done when the “BUGs” occur since these should not happen once the code is debugged.

To make the code clearer, “Error_” was added to all the error and “BUG_” was added to all the diagnostic error Translator status enumeration names. By the way, each of theses codes is only used in one place so that it is easy to figure where an error occurs (if one “stack empty” was used instead of five unique codes, then if one occurs, there would be confusion which place in the code it came from).

The code now checks the number of arguments for internal functions and ibcp_0.1.7-src.zip has been uploaded at Sourceforge IBCP Project along with the binary for the program. Next handling assignment statements and the concept of references...

Translator – Internal Functions (Implementation)

The number of arguments member needs to be added to the TableEntry structure and values to the table entries along with an access function to this member in the Table class. New entries are needed for the MID$, ASC and INSTR functions with the appropriate number of arguments. A new multiple entry flag is needed to the first entry for each of these functions to indicate that more entries exist with different number of arguments.

The check for the number of arguments for internal functions will be added to the closing parentheses token processing in the array or function section. If the token popped off of the hold stack is an internal function and the number of operands doesn't match the number of arguments in the table, then if the multiple entry flag is set it needs to search for another entry of the function that matches the number of operands. If found, the the token's index will be changed to the new entry's index. Otherwise a “wrong number of arguments” error will occur.

A new table search function is needed that will search for a function with a specific number of arguments. The index of the first entry for the function will be specified assuming that it has already been checked for the number of arguments (this is the index the Parser originally returned). The search will proceed with the entry after the index and continue until the end of the internal functions (when an entry with a NULL name if found). The additional entries must be in the same internal functions section (words with parentheses).

Translator – Internal Functions

The Translator is unable to check the number of operands for arrays or functions without more information – information that will be contained in the Dictionary, so this checking will be left for the Encoder. However, the number of arguments for internal functions are fixed and can be checked within the Translator.

The number of arguments cannot be checked with knowing the number, therefore a new number of arguments member needs to be added to the Table entries. However, some internal functions may have more than one form. Several of these types or functions are currently planned, which include MID$, ASC and INSTR.

This means that there will need to be multiple Table entries, like for the existing Sub_Code and Neq_Code for the minus operator. For example with MID$, there will be Mid2_Code and Mid3_Code. Initially the Parser will set the code to Mid2_Code since that will be the first entry found in the table.

When the MID$ token is about to be added to the output list, the Translator will have the number of operands that needs to be checked. If the number matches what's in the table, nothing further needs to be done. If the number doesn't match, the Translator either needs to return an error or go looking for another Table entry with the correct number of operands found. Therefore, there needs to be a flag in the Table that indicates that there are more entries. If this flag is set then the Translator needs to look for it.

Translator – Arrays and Functions (Release)

The SimpleStack class is in the new source file stack.h. A simple stack test program was written to test the SimpleStack class, especially the automatic expanding feature of SimpleStack. There is no VIDE2 project file for with program, it was compiled with the command line “g++ test/test_stack2.cpp -o teststack2.exe” via MSYS.

Several test expressions were added for testing arrays and functions including multiple levels of  identifiers with parentheses, defined functions and internal functions. At this point there is no difference between arrays and user functions – both are identifiers with parentheses. The last several expressions are ones that were used to test error conditions (these actually caused problems during testing).

A change was made to how the test output is generated specifically related to internal operator codes. Right now there is only one of these, the Negate code. Originally the name string was set to “Neg” to distinguish between “-” (binary operator subtraction) and “-” (unary operator negation) in the test output. However, the name will eventually need to be “-” for the Recreator. There will be many more of these operators. Therefore, name2 will now be used for the test output and name will be for the actual output. For test output, if name2 is not set (NULL) like it is for all operators, name will be used for the test output, otherwise name2 will be used.

The code now handles arrays and functions and ibcp_0.1.6-src.zip has been uploaded at Sourceforge IBCP Project along with the binary for the program (the test stack program is not included). Next the checking for the correct number of arguments for internal functions will be added to the Translator...

Translator – Arrays and Functions (Implementation)

For the new counter stack in the Translator class, the SimpleStack class was implemented. This class is a very simple stripped down version of the List class implementation that uses simple allocated array (expanded as needed) instead of using a double linked list. The template contains arguments for the initial size of the array and the size the array is increased when it gets filled (both arguments default to 10).  A key feature of this class is that the top() function can be used to manipulate the item on top of the stack (in this case a comma counter) directly.

When an array or function token is encountered (any token that has a parentheses), it is pushed onto the hold stack. The Translator state is left set to Operand since the next token must be an operand (or optional unary operator). The pending parentheses check needs to be made before the token that has a parentheses is handled.

For comma handling, a 0 is pushed onto the counter stack for an open parentheses token. If a comma token is encountered and the counter stack is either empty or the top of the counter stack is 0, then an “unexpected comma” error occurs. For a token that has a parentheses, a 1 is pushed onto the counter stack. For each comma token encountered, the counter on top of the stack is incremented.

Now when a closing parentheses token is encountered, if the counter stack is empty, then a “missing opening parentheses” occurs. The top counter is popped off the stack. If this counter value is zero, then there was an opening parentheses, so an open parentheses token is expected on top of the hold stack. The pending parentheses token pointer only needs to be set for closing parentheses tokens.

When the value popped from the top of the counter stack is not zero, it contains the number of operands for an array or function, which is expected on top of the hold stack (specifically a token that has a parentheses). For now, the operands are just popped from the done stack. The array or function token is popped from the hold stack, added to the output list and pushed on the done stack.

Other changes needed was the precedence check when emptying the hold stack to work with tokens that don't have table entries (array and function tokens); at the EOL processing, adding a token has parentheses to the missing closing parentheses check; and emptying the counter stack in the clean up routine.

Translator – Arrays and Functions (Design Issues)

The first issue is that the number of operands counter cannot be a simple variable because operands of arrays and functions can be other arrays and functions, so there needs to be a counter variable for each nested array or function. Therefore a counter stack is needed for multiple nested levels of arrays or functions. Using the existing List class for this simple counter stack would be overkill, so a new simple stack class will be developed with a similar interface (so it would be easy to switch from one to the other for stacks).

The second issue is when an array or function token is pushed on the stack. The Paren and DefFuncP token types do not have a table index and therefore a precedence can't be obtained from a table entry. To resolve this, a new precedence() function is needed in the Token class, which will work the same as the is_operator() and has_paren() functions using a static precedence array. A new precedence() function will be added to the Table class with a Token pointer argument that will use the new Token precedence() function.

The values in the static precedence array within the Token class will be initialized to -1 if the token type has an associated table entry, otherwise the value is set to the desired precedence. For the Command, Operator and IntFuncP token types that have table entries, the token precedence will be set to -1 and for the DefFuncP and Paren token types, the precedence will be set to 2 (the same as the open parentheses precedence). The other token types won't be pushed to the hold stack so no value is needed at the moment.